Overview
Welcome back to your mission control
Monitor key signals from every agent-driven workflow, check live performance, and access the latest automations in one action-oriented view.
Active Applications
4
Each application ships with shared AI infrastructure and tailored workflows.
Automation Coverage
82%
Track how much of your operations are fully automated by autonomous agents.
Quick Actions
- Launch workflow builder New
- Invite a teammate 6 pending
- Review AI transcripts 12 today
Record Snapshot
Candidate Profile · Jordan Fox
Dynamic record view powered by PHP endpoints like GET /api/candidates/{id} and GET /api/candidates/{id}/timeline.
Smart Summary
LLM overview of candidate readiness synthesized from interviews, assignments, and recruiter notes.
Highlights: Excels in customer onboarding, shipped AI-driven knowledge base migration at BluePeak Logistics.
Concerns: Requires enablement on SOC 2 evidence workflows.
Recommendation: Proceed to offer creation pending compensation alignment.
Responses cached via candidate_summaries table.
Timeline
-
🎤
Interview · Solution Assessment
2025-10-09 14:00 · Panel: Taylor Reed, Marco Alvarez
AI transcription stored in
interview_transcripts. Score 4.6 / 5. -
📝
Assignment Submitted
2025-10-06 18:42 · Automation Playbook Draft
Reviewers added to
candidate_documents. Pending final redline. -
💬
Recruiter Touchpoint
2025-10-04 10:18 · Slack connect
Logged in
activitiesviaPOST /api/activities.
Extend with pagination from timeline_events table.
Related Panels
`candidate_links` tableOpen Roles
Customer Experience Lead
Hiring Manager: Marco Alvarez
References
Alex Kim · Former Director
Status: Pending outreach
Offers
Draft Offer Package
Comp: $145K base · Bonus 12%
Linked Accounts
BluePeak Logistics
Sourced via referral program
Enterprise CRM Agent
Sales Agent (CRM)
Deploy autonomous pipeline management with real-time qualification, objection handling, and dynamic forecasting across your revenue teams. Connect to PHP REST endpoints (e.g. GET /api/crm/leads, POST /api/crm/opportunities) that orchestrate the leads, accounts, opportunities, quotes, and activities tables.
Lead → MQL
45.2%
▲ 8.3% vs last 30 days
MQL → SQL
68.5%
▲ 12.1% vs last quarter
Win Rate
32.1%
▼ 2.3% vs target
Avg Deal Size
$47.5K
▲ $4.2K vs last month
Pipeline Kanban
Visualize every opportunity from Prospecting to Won with AI recommendations on acceleration and risk. Data loads from GET /api/crm/pipeline backed by the opportunities table.
Prospecting
14-
Acme Data Cloud
$84,500 · probability 20%
AI note: awaiting security questionnaire.
-
Brightline Health
$36,000 · probability 25%
Next step: schedule discovery.
Qualification
11-
Nimbus AI
$58,750 · probability 45%
AI suggests looping legal for DPA draft.
-
Lumen Solar
$120,000 · probability 40%
BANT complete · timeline Q4.
Proposal
9-
Helix Ventures
$240,000 · probability 65%
Quote v2 pending CFO sign-off.
-
Orion Telecom
$180,000 · probability 55%
AI recommends sharing ROI calculator.
Won
9-
Atlas Robotics
$320,000 · probability 100%
Contract synced to billing · kickoff booked.
-
BluePeak Logistics
$95,000 · probability 100%
Implementation handoff complete.
Lead Workspace
Manage inbound demand with AI scoring from the leads table. Filters map to PHP query params on GET /api/crm/leads.
| Lead | Source | Status | Score | Owner | Last Interaction | |
|---|---|---|---|---|---|---|
|
Hannah Cole Budget: $60K · Timeline: 45 days |
Web | Contacted | 78 | Alex Rivera | AI note: waiting on stakeholder matrix. | |
|
Orbit Labs Pain points: scaling onboarding, manual quotes. |
Referral | Qualified | 91 | Priya Chen | Meeting scheduled · Oct 18 | |
|
Jameson Freight Budget TBD · Timeline: 90 days |
New | 42 | Unassigned | AI follow-up queued · Oct 13, 09:00 |
GDPR consent stored in leads.gdpr_consent; ensure PHP middleware enforces consent logging.
Account Intelligence
Surface company hierarchies, expansion potential, and health scores from the accounts table.
Top Expansion
Lumen Solar
ARR $420K · Upsell +24%
At Risk
3 accounts
Usage dropped 30% WoW
Enterprise Coverage
12 workspaces
Assignments balanced via territory rules.
| Account | Industry | Tier | Annual Revenue | Health | Owner | |
|---|---|---|---|---|---|---|
|
Atlas Robotics Employees: 950 |
Manufacturing | Enterprise | $48.0M | Healthy | Priya Chen | |
|
Helix Ventures Employees: 120 |
Finance | Mid-Market | $12.4M | Monitor | Taylor Reed | |
|
Nimbus AI Employees: 300 |
Software | Mid-Market | $27.0M | Risk | Alex Rivera |
Opportunity Tracker
Track deal velocity, projected revenue, and AI strategy briefs from opportunities.
Pipeline
$2.4M
▲ $380K vs Aug
Avg Cycle
28 days
▼ 5 days
Meetings
43
▲ 12 booked
Quotes Sent
87
76% accepted
| Opportunity | Account | Stage | Amount | Probability | Close Date | Strategy | |
|---|---|---|---|---|---|---|---|
| Atlas Robotics Renewal | Atlas Robotics | Negotiation | $320,000 | 80% | Oct 22 | Loop CFO in on ROI deck · confirm security summary. | |
| Nimbus AI Expansion | Nimbus AI | Proposal | $240,000 | 55% | Nov 05 | Bring technical architect; address AI governance questions. | |
| Brightline Health Pilot | Brightline Health | Qualification | $36,000 | 35% | Dec 10 | Send compliance checklist · prepare stakeholder map. |
Quote Generator
Create and manage proposals stored in the quotes table. PHP endpoint placeholder: POST /api/crm/quotes.
Draft
18
Avg approval 4h
Sent
52
Awaiting client
Accepted
76%
+9% MoM
| Quote | Opportunity | Total | Currency | Validity | Status | |
|---|---|---|---|---|---|---|
| Q-2025-084 | Nimbus AI Expansion | $240,000 | USD | Oct 30 | Sent | |
| Q-2025-079 | Atlas Robotics Renewal | $320,000 | USD | Oct 20 | Accepted | |
| Q-2025-071 | Brightline Health Pilot | $36,000 | USD | Nov 05 | Draft |
Meeting Scheduler
Sync calendars and capture AI notes saved as activities rows (type='meeting').
Next Available
Oct 15 · 10:30 AM
Auto-generated from synced calendars.
Average Prep
6 min
AI briefs cut prep time by 72%.
Attendance
94%
+5% with auto-reminders
Today
-
Discovery · Orbit Labs
09:30 · Alex Rivera · Zoom
Agenda: Persona alignment · integration scoping
-
Renewal Strategy · Atlas Robotics
13:00 · Priya Chen · Google Meet
AI note: share product roadmap highlights
Tomorrow
-
Proposal Review · Nimbus AI
11:00 · Taylor Reed · Teams
AI note: confirm security posture deck
AI Agenda Assistant
Generate tailored agendas by posting to POST /api/crm/meetings/{id}/agenda with participant list.
Meeting Recap Feed
- 09:20 · Orbit Labs · Summary posted to Slack.
- 09:25 · Orbit Labs · Follow-up email drafted.
- 13:45 · Atlas Robotics · Action items synced to Jira.
Unified Activity Timeline
View omnichannel engagement stored in the activities table with PHP endpoint GET /api/crm/activities/feed.
Call · Nimbus AI · Alex Rivera
10:05 · Outcome: Qualified · Duration 18m
AI summary pushed to CRM. Action item: send pricing guardrails.
Email · Orbit Labs · Priya Chen
09:42 · Auto-drafted follow-up · Sent via Gmail API
CTA: confirm security review availability.
Meeting · Atlas Robotics Renewal
Yesterday · Notes synced to Notion
Summary: contract extended 24 months · upsell success.
AI Chat · Jameson Freight
Oct 11 · WhatsApp channel
Bot recommended shipping automation playbook · human follow-up scheduled.
Autonomous Support Desk
Customer Support Agent
Deliver intelligent triage, deflection analytics, and SLA-aware routing with AI copilots assisting every interaction. Front-end panels expect PHP endpoints such as GET /api/support/tickets, POST /api/support/queues/{id}/assign, and GET /api/support/kpis mapping to the `tickets`, `queues`, `sla_policies`, and `knowledge_articles` tables.
Deflection
67.3%
▲ 9.5% vs last month
First Response Time
2m 14s
▼ 34s vs target
Mean Time To Resolve
18m 42s
▼ 4m 18s vs last week
SLA Met
94.2%
▲ 3.8% vs quarter start
Omnichannel Live Inbox
Monitor live chats, emails, and WhatsApp threads with AI-suggested replies and disposition logging. Bind to GET /api/support/inbox.
-
TCK-1384 · Password reset
LiveWhatsApp · Hannah Cole · Priority: Medium
AI suggestion: send secure reset link and confirm MFA status.
-
CS-9876 · Order status
Awaiting ReplyEmail · Marcus Lee · Priority: Low
AI drafted response ready. SLA: 12m remaining.
Ticket Command Center
Full lifecycle visibility for records stored in the tickets table.
| Ticket | Priority | Status | Queue | SLA | Owner | |
|---|---|---|---|---|---|---|
|
TCK-1384 · Password reset failing SSO Contact: Hannah Cole · Created 14:22 |
High | Open | Account Access | At Risk · 4m remaining | Alex Rivera | |
|
CS-9876 · Order shipment delay Contact: Marcus Lee · Created 09:05 |
Medium | Pending | Logistics | Within · 1h 12m | Priya Chen | |
|
HD-5521 · Refund request Contact: Sara Lopez · Created yesterday |
Low | Resolved | Billing | Completed · 6h early | Zoe Patel |
Queues & Routing
Configure capacity and automations stored in the queues table.
Inbound Volume
486 tickets
▼ 12% vs last week
Auto-Routed
91%
▲ 6% with AI triage
Breached
3 queues
Review staffing
| Queue | Type | Active Tickets | Members | Routing Rules | |
|---|---|---|---|---|---|
| Global Tier 1 | by_team | 72 | Alex, Priya, Zoe, Taylor | Round-robin · business hours only | |
| Enterprise Escalations | by_skill | 18 | Taylor, Amir | Route when severity ≥ major | |
| Billing Operations | by_region | 32 | EMEA Finance Squad | Route by geo & account tier |
SLA Policies & Timers
Manage policy definitions (`sla_policies`) and live timers (`sla_timers`).
Policies
8 active
Channel-specific coverage
Breaches Today
5
Notify incident team
Paused Timers
11
Awaiting customer reply
| Policy | Priority | Channel | First Response | Resolution | Business Hours | |
|---|---|---|---|---|---|---|
| Critical Incident | Urgent | All | 5 min | 30 min | No | |
| Standard Support | Medium | 1 hr | 8 hr | Yes | ||
| Premium Voice | High | Voice | 2 min | 45 min | No |
Live SLA Timers
-
4m left
TCK-1384 · First Response
Target 15:04 · Paused 0m
-
On Track
CS-9876 · Resolution
Target 17:00 · Paused 8m
Knowledge Base
Manage `knowledge_articles` with AI-assisted drafting from resolved tickets.
Published
312
▲ 24 new this month
Deflection
67.3%
↑ with AI article suggestions
Update Needed
18
Auto-flagged over 90 days
| Article | Category | Helpful | Deflections | Status | |
|---|---|---|---|---|---|
|
How to reset SSO password Version 3 · Updated Oct 10 |
Account Access | 134 / 6 | 267 | Published | |
|
Refund processing timelines Version 1 · Draft |
Billing | 28 / 2 | 54 | Draft | |
|
Troubleshooting webhook failures Version 2 · Review needed |
Integrations | 89 / 11 | 102 | Update |
AI Draft Preview
Generated from ticket TCK-1384 after resolution.
Summary
Steps 1. Verify user identity via MFA. 2. Use admin console → Security → Reset password. 3. Send magic link; advise user to update within 24 hours.
Support Analytics
Blend operational metrics with AI predictions. Connect to GET /api/support/reports.
CSAT
4.6 / 5.0
▲ 0.3
First Contact Resolution
78.9%
▲ 8.1%
Reopen Rate
4.2%
▼ 1.1%
Backlog
23 tickets
▼ 12
Channel Performance
FRT 3m 12s · CSAT 4.7
Chat
FRT 46s · CSAT 4.5
FRT 1m 08s · CSAT 4.4
Forecast
AI projects backlog clearance by Oct 19 with current staffing.
- Daily Inflow: 68 tickets · ▼ 4%
- Required Agents: 12 · coverage 105%
- Predicted Breaches: 6 (focus on Enterprise Escalations)
Recruiting Intelligence
AI Interviewer (ATS)
Run structured, bias-aware hiring pipelines with autonomous interview orchestration, confidence-scored scorecards, and compliant offer workflows.
Live Insights
Time-to-Hire
18.4 days
Down 5.6 days vs last quarter
Candidate NPS
8.7 / 10
Up 1.2 points in 30 days
Apply → Screen
34.2%
Up 6.8% with AI parsing
Screen → Offer
28.5%
Up 4.1% after rubric alignment
Jobs & Requisitions
Build requisitions, auto-generate job descriptions, and publish across job boards with guardrails for compliance and compensation bands.
| Title | Department | Location | Type | Status | Hiring Manager | Candidates | Updated | |
|---|---|---|---|---|---|---|---|---|
| Senior Product Manager | Product | Remote - North America | Full Time | Open | Taylor Reed | 32 | Oct 11 | |
| Staff Backend Engineer | Engineering | Hybrid - San Francisco | Full Time | On Hold | Priya Chen | 14 | Oct 9 | |
| Customer Success Manager | Customer Success | Remote - EMEA | Full Time | Open | Jordan Lee | 21 | Oct 7 |
Candidate Workspace
Track talent signals, orchestrate outreach, and sync enriched profiles directly from the unified candidate graph.
Total Candidates
1,248
▲ 82 new this week
AI Fit Score (avg)
78
Weighted by target job competency matrix.
Diversity Funnel
41%
▲ 3% vs last quarter
Automation Coverage
68%
Sequences live for nurture, reminders, and feedback loops.
| Candidate | Current Role | Stage | Fit Score | Source | Applied | Last Touch | |
|---|---|---|---|---|---|---|---|
|
Avery Collins Skills: Product Strategy · SQL · AI Co-Pilot |
Senior Product Manager · Compass AI | Interview | 86 | Sep 27 | Oct 11 · AI Note Added | ||
|
Noah Patel Skills: Backend · Python · Microservices |
Staff Platform Engineer · CloudWave | Phone Screen | 74 | Referral | Sep 30 | Oct 10 · Reminder Sent | |
|
Layla Chen Skills: Customer Success · Salesforce · CS Ops |
Enterprise CSM · FlowDesk | Offer | 92 | Website | Oct 4 | Oct 11 · Offer Drafted | |
|
Isaiah Brooks Skills: Support Ops · Zendesk · QA |
Support Lead · PulseCX | Rejected | 58 | Job Board | Sep 19 | Oct 6 · Feedback Sent |
Candidate Spotlight
Avery Collins
Product Leadership · Remote (USA)
Experience: 6.5 years · Compass AI · Zillow
Fit Highlights: AI roadmap ownership, PLG analytics, GTM launch playbooks.
Next Step: Panel interview on Oct 14 · Scorecard assigned to AI Evaluator + Taylor Reed.
Smart Segments
- Ready for AI Screen 58
- Passive Talent Nurture 142
- Requires Feedback 19
- Offer Looming (7d) 8
Pipeline Health
Monitor conversion rates across stages and auto-trigger nudges when candidates stall.
Applied
512
Phone Screen
176
Interview
98
Offer
28
Hired
24
Drop-Off
53%
Stage Conversion
Risk & Alerts
-
Design Engineer pipeline trending low
Drop-off 42% at phone screen. Recommend enabling AI async interview.
-
Interview panel bandwidth risk
Only 6 available panelists next week. Auto-suggest reschedules.
-
Diversity funnel improving
URM representation up 3% at final stage vs prior cohort.
Interview Operations
Coordinate human and AI interviews, manage transcripts, and sync outcomes to scorecards.
Today's Schedule
Noah Patel · AI Technical Screen
09:00 AMPanel: Priya Chen (Human) · AI Co-pilot
Avery Collins · Product Loop
11:30 AMPanel: Taylor Reed, Sara Malik · Hybrid
Mia Ramirez · Support Simulation
02:00 PMMode: Async AI Interview · Duration 25 min
Panel Availability
Sync interviewer calendars via `/php/integrations/calendar` webhooks to avoid conflicts.
Scorecards & Feedback
Standardize evaluation rubrics, surface AI highlights, and route bias checks before final decisions.
Avg Rating
8.6 / 10
Weighted by competency importance.
Strong Yes Rate
41%
Up 7% since calibrating rubric.
Bias Flags
3
Requires review in compliance queue.
| Candidate | Interview | Evaluator | Overall | Recommendation | Bias Flags | |
|---|---|---|---|---|---|---|
| Avery Collins | Product Loop | Sara Malik (Human) | 9 | Strong Yes | 0 | |
| Noah Patel | AI Technical Screen | AI Evaluator | 8 | Yes | 1 · Tone Drift | |
| Mia Ramirez | Support Simulation | Jordan Lee (Human) | 6 | Maybe | 2 · Bias review |
Offers & Approvals
Automate offer letter generation, capture approvals, and sync to HRIS once accepted.
Draft
6
Sent
5
Accepted
4
Expired
1
| Candidate | Title | Salary | Equity | Start Date | Status | |
|---|---|---|---|---|---|---|
| Avery Collins | Senior Product Manager | $185k · USD | 0.08% equity | Nov 18 | Accepted | |
| Noah Patel | Staff Backend Engineer | $215k · USD | 0.12% equity | Dec 2 | Sent | |
| Lia Nguyen | Customer Success Manager | €86k · EUR | €5k bonus | Nov 25 | Expired |
Reports & Analytics
Track KPIs across the ATS funnel. Connect to PHP reporting endpoints or BI exports.
AI Interviewer KPIs
Time-to-Hire
18.4 days
↓ 5.6 days
Candidate NPS
8.7 / 10
↑ 1.2
Apply → Screen
34.2%
↑ 6.8%
Screen → Offer
28.5%
↑ 4.1%
Source & Diversity
↑ 3% representation
30-Day Trendline
Time-to-hire trending down as async AI interviews increase throughput. Forecast shows 16.8 days by November 1.
Upskilling & Coaching
AI Mock Interviewer
Deliver hyper-personalized practice interviews, skill dashboards, and instructor collaboration for career-readiness.
Practice Overview
Monitor student performance trends, job-readiness KPIs, and AI-powered feedback quality across the mock interview program.
Completion Rate
82.4%
↑ 6.2% vs last cohort
Avg Improvement
+18.7 pts
↑ 3.4 pts post coaching
Job Offers
34.2%
↑ 8.1% vs baseline
Engagement
142 min/wk
↑ 23 min vs last month
Connect to PHP `/php/mock/dashboard` for live KPI refresh from `students`, `practice_sessions`, and `badges` tables.
Program Automation
- Scenario Selection: `GET /php/mock/question-banks` filtered by role.
- Mock Interview: Start async session via `/php/mock/sessions/start` → connects to FastAPI for live coaching.
- Feedback Generation: PHP job `/php/mock/assessments/summarize` writes rubric scores to `assessments`.
- Portfolio Showcase: `/php/mock/showcase/publish` generates shareable badge pages.
Student Directory
Manage learner rosters, consent preferences, and target roles to personalize mock interview tracks.
| Student | Program | Target Roles | Graduation | Consent | Sessions | |
|---|---|---|---|---|---|---|
| Jordan Kim | Bootcamp | Software Engineer, Backend | 2025 | Showcase | 6 | |
| Amina Farouk | University | Product Manager | 2026 | Data Only | 4 | |
| Diego Alvarez | Self Study | Data Scientist, ML Engineer | 2025 | Pending | 5 |
CRM Sync
Connect to `/php/mock/students/sync` to pull latest contact records from CRM into the `students` table.
Onboarding Tasks
- Assign practice track via `/php/mock/assignments`.
- Invite coach using `/php/mock/coaches/:id/assign`.
- Enable showcase consent toggle.
Store student docs (CV, portfolio) at `students.cv_url` & `students.portfolio_url`. Surface downloads in learner profile modals.
Practice Sessions
Track every AI-led practice with transcripts, rubric scores, and readiness signals.
| Student | Role | Company Scenario | Mode | Duration | Score | Status | |
|---|---|---|---|---|---|---|---|
| Jordan Kim | Software Engineer | Stripe – Payment Reliability | Video | 42 min | 87 | Completed | |
| Amina Farouk | Product Manager | Notion – Collaboration Vision | Voice | 35 min | 91 | Scheduled | |
| Diego Alvarez | Data Scientist | Airbnb – Search Ranking | Text | 28 min | 79 | In Progress |
Assessment Rubrics
Review structured scoring output, filler word analysis, and personalized recommendations from each mock interview.
| Student | Session | Overall Score | Filler Words | Speaking Pace | Highlights | |
|---|---|---|---|---|---|---|
| Jordan Kim | Stripe – Payment Reliability | 87 | 14 | 152 wpm | Communication 8 · Structure 7 · Technical 9 | |
| Amina Farouk | Notion – Collaboration Vision | 91 | 9 | 138 wpm | Behavioral 9 · Product Thinking 9 · Narrative 8 | |
| Diego Alvarez | Airbnb – Search Ranking | 79 | 22 | 164 wpm | Needs concise storytelling · Highlight metrics |
Evidence Timeline
`assessments.evidence` stores key timestamped clips. Render inside session viewer using `/php/mock/assessments/:id/evidence`.
Kick off auto-coaching by calling `/php/mock/assessments/summarize` once transcripts finish processing in FastAPI.
Assignments
Align students to targeted drills with automated reminders and instructor feedback loops.
| Student | Role | Company Scenario | Mode | Due Date | Status | |
|---|---|---|---|---|---|---|
| Jordan Kim | Software Engineer | Stripe – Payment Reliability | Video | Oct 20 | Completed | |
| Amina Farouk | Product Manager | Notion – Collaboration Vision | Voice | Oct 18 | Scheduled | |
| Diego Alvarez | Data Scientist | Airbnb – Search Ranking | Text | Oct 16 | In Progress |
Cohort Intelligence
Organize learners into programs, compare readiness, and surface at-risk students for instructor outreach.
| Cohort | Instructor | Learners | Avg Score | At Risk | Next Session | |
|---|---|---|---|---|---|---|
| SE Q4 Intensive | Priya Chen | 48 | 86 | 7 | Oct 14 · 10:00 AM | |
| PM Launchpad | Marcus Lee | 36 | 84 | 5 | Oct 15 · 4:00 PM | |
| Data Science Residency | Jasmine Reed | 28 | 88 | 3 | Oct 16 · 1:00 PM |
Cohort Slack
Sync announcements via `/php/mock/cohorts/:id/announcements` and mirror into Slack channels.
Risk Radar
Surface students with `assessments.overall_score < 75` or `practice_sessions.status = 'overdue'` for intervention.
Use `/php/mock/cohorts/export` to populate BI dashboards tracking cohort velocity, placement rate, and mentor load.
Question Bank
Maintain structured prompts, sample answers, and difficulty tags for AI-led practice scenarios.
| Prompt | Role | Type | Difficulty | Sample Answer | |
|---|---|---|---|---|---|
| Tell me about a time you improved system reliability. | Software Engineer | Behavioral | Medium | Highlights metrics, root cause, stakeholders. | |
| Design a feature prioritization framework for roadmap planning. | Product Manager | Case Study | Hard | Discuss RICE model, stakeholder alignment, validation. | |
| Walk through your process for evaluating experiment results. | Data Scientist | Technical | Medium | Covers hypothesis, metrics, statistical significance. |
Scenario Templates
`question_banks.artifacts` can store slides, whiteboard prompts, or datasets. Connect via `/php/files/:id`.
Leverage `/php/mock/question-banks/generate` to auto-create variations using GPT while keeping instructor approval in the loop.
Rubric Builder
Design and maintain scoring frameworks with weighted criteria and defined scales tailored to each interview role.
| Rubric | Role | Criteria | Scale | Weighting | |
|---|---|---|---|---|---|
| SE Technical Rubric | Software Engineer | Communication, System Design, Coding | 1-10 | 0.3 / 0.4 / 0.3 | |
| PM Narrative Rubric | Product Manager | Storytelling, Product Strategy, Metrics | 1-5 | 0.25 / 0.5 / 0.25 | |
| DS Experimentation Rubric | Data Scientist | Hypothesis, Analysis, Communication | 1-5 | Balanced |
Criteria JSON
`rubrics.criteria` holds structured arrays like [{ name, weight, scale }]. Use `/php/mock/rubrics/:id` to hydrate builders.
Trigger rubric versioning via `/php/mock/rubrics/:id/version` to preserve historical scoring logic.
Badges & Certificates
Celebrate learner milestones with verifiable credentials, badge evidence, and shareable portfolio links.
| Student | Badge | Evidence Sessions | Issued | Certificate | |
|---|---|---|---|---|---|
| Jordan Kim | Technical Proficiency | Session IDs: 4112, 4119 | Oct 10 | View PDF | |
| Amina Farouk | Communication Expert | Session IDs: 5201, 5202, 5203 | Oct 8 | View PDF | |
| Diego Alvarez | 10 Sessions Milestone | Session IDs: 6120-6129 | Oct 5 | View PDF |
Portfolio Pages
Use `/php/mock/badges/:id/share` to generate unique showcase URLs from `badges.certificate_url`.
Sync earned badges to LinkedIn via `/php/integrations/linkedin` once student opts into showcase consent.
Analytics
Track cohort readiness, rubric improvements, and interview throughput to continuously optimize coaching.
Avg Improvement
+18.7 pts
↑ 3.4 vs prior cohort
Engagement
142 min
↑ 23 min weekly
Job Offers
34.2%
↑ 8.1% conversion
NPS
9.1 / 10
↑ 0.7 last 30 days
Top Improvements
- Jordan Kim: +22 pts technical · +18 pts communication
- Amina Farouk: +17 pts narrative · +11 pts structure
- Diego Alvarez: +14 pts problem-solving · +9 pts storytelling
Leaderboard combines `assessments.rubric` aggregates with `practice_sessions.duration_minutes` to gauge improvement velocity.
Agent Configurations
Design and version AI agents across workspaces. Mirrors the `agents` and `agent_versions` tables; connect to PHP endpoints like `GET /api/agents`, `POST /api/agents`, and `PUT /api/agents/{id}/versions`.
Active Agents
18
Rows counted from `agents` where status = 'active'.
Drafts Awaiting Review
6
`status = 'draft'` and `current_version_id IS NULL`.
Channels Connected
11
Aggregate distinct values from `agent_versions.channels` JSON.
Last 7 Days Updates
9
`agent_versions.created_at >= CURRENT_DATE - INTERVAL '7 days'`.
| Agent | Workspace | Type | Current Version | Status | Channels | Updated | |
|---|---|---|---|---|---|---|---|
| Outbound Revenue Coach | Sales APAC | Sales | v7 (claude-3-5-sonnet) | Active | Web, Email, WhatsApp | 2025-10-07 | |
| Support Concierge | Support Americas | Support | v5 (gpt-4o-mini) | Active | Web, Voice | 2025-09-30 | |
| Talent Screener | Global HQ | ATS | v3 (claude-3-haiku) | Draft | Web | 2025-09-14 |
Version History
v7 · claude-3-5-sonnet
LivePrompt tuned for enterprise outbound cadence. Tools: `crm_create_lead`, `calendar_booking`.
Created 2025-10-01 by Priya Chen
v6 · claude-3-opus
Rollback optionIntroduced voice channel, temperature 0.65. Guardrails: `pii_scrubber`.
Created 2025-09-10 by Taylor Reed
Deploy Agent
Trigger rollout to production. Uses `POST /api/agents/{id}/deploy` (PHP) with callbacks to the Python microservice.
Agent Blueprint
Database Mapping
Updates persist to `agents` (columns: `name`, `type`, `status`, `workspace_id`). Setting current_version_id auto-hydrates deployment widget.
Persisted to `agent_versions.channels` JSON array.
Create Version
Stored in `agent_versions.system_prompt` (TEXT). Last modified timestamp updates `agent_versions.created_at`.
FastAPI Integration Placeholder
On save, call `/ai/chat/stream` in dry-run mode to validate guardrails and response quality before enabling is_active.
Change Log
v7 promoted to production
2025-10-07 · by Priya Chen · Endpoint: `POST /api/agents/{id}/deploy`
Channels extended to WhatsApp. Guardrails added: tone_monitor.
v6 rollback prepared
2025-09-12 · by Taylor Reed · Endpoint: `PUT /api/agents/{id}/versions/{version_id}`
Introduced voice channel with monitoring toggled on.
Draft saved
2025-09-01 · by Morgan Ortiz · Endpoint: `POST /api/agents`
Initial agent shell created for outbound pilot workspace.
Prompt Templates
Curate reusable prompt assets for agents. Stored in `agent_versions.system_prompt` and auxiliary template stores. Surface via PHP endpoints or the FastAPI prompt library.
Prompt Library
Sales Discovery v2
Targets enterprise SaaS. Includes persona traits, qualification steps (BANT), and fallback objections.
Linked Agents: Outbound Revenue Coach (v7)
Support Troubleshooter v4
Step-by-step troubleshooting flows. Integrates with knowledge articles and automations.
Linked Agents: Support Concierge (v5)
Prompt Preview
System Message
You are an autonomous sales assistant for DoBeyond. Prioritize enterprise SaaS buyers. - Collect budget, timeline, primary stakeholders - Offer relevant case studies from `crm_case_studies` - Book follow-up using `calendar_booking` Tone: consultative, confident, value-led.
Backend Placeholder
Connect to PHP `PUT /api/agents/{id}/versions` to persist edits. For runtime evaluation, call FastAPI `/ai/chat/stream` with this prompt as the system message.
Skills & Tools
Attach external capabilities to agent versions. Stored inside `agent_versions.tools` JSONB array (e.g., `['rag_search', 'calendar_booking']`).
| Tool | Description | Linked Agents | Last Sync | |
|---|---|---|---|---|
| rag_search | Vector retrieval service with pgvector for proprietary documents. | Outbound Revenue Coach, Support Concierge | 2025-10-08 | |
| calendar_booking | Integrates with Google Workspace / Outlook for live scheduling. | Outbound Revenue Coach | 2025-09-27 | |
| crm_create_lead | Writes qualified leads into HubSpot/Salesforce via PHP webhook. | Outbound Revenue Coach | 2025-09-02 |
Channels & Guardrails
Channels
Web, WhatsApp, Voice, Email
Guardrails
PII scrubbing, tone moderation, escalation triggers.
Python Service
FastAPI `/ai/chat/stream` handles execution with SSE.
A/B Testing
Compare agent variants and pick winners. Reflects the `ab_tests` table with endpoints like `GET /api/agents/{id}/ab-tests` and FastAPI evaluation callbacks.
| Test Name | Variant | Status | Start Date | End Date | Winning Variant | Results | Actions |
|---|---|---|---|---|---|---|---|
| Sales Pitch Optimization | Variant A | Active | 2025-10-15 | 2025-11-15 | Variant B | Pending | |
| Customer Support Chatbot | Variant B | Pending | 2025-10-20 | 2025-11-20 | Variant C | Pending | |
| AI Interviewer Accuracy | Variant C | Draft | 2025-10-25 | 2025-11-25 | Variant A | Pending |
Live Inbox
Monitor all active conversations streaming in from web, WhatsApp, voice, and email channels. Backed by the `conversations` and `messages` tables with endpoints such as `GET /api/conversations` and `POST /api/conversations/{id}/messages`.
API Hook
Use `GET /api/conversations?status=active` every 5 seconds or subscribe to a websocket topic for new messages. On selection, hydrate message pane via `GET /api/conversations/{id}/messages`.
Enterprise Renewal · Conversation #C-4821
Channel: Web · Status: Active · Outcome: Pending
User · 08:23
We're renewing for Q4 but want to confirm the automation limits for voice channels.
Assistant · 08:24 · Tokens: 148 · Latency: 312ms
Automation coverage includes 100 voice minutes per seat. I can route a pricing breakdown or loop in billing if you'd like.
Tool · crm_create_lead
Lead synced to HubSpot. Confidence 0.94. Intent: Expansion.
Stream Placeholder
Use Server-Sent Events or WebSockets to append new entries to this timeline in real time. Persist each turn in the `messages` table with metadata such as `intent`, `sentiment`, `token_count`, and `latency_ms`.
All Conversations
Search historical transcripts across every channel. Mirrors the `conversations` table and supports pagination via `GET /api/conversations`.
API placeholder: call `/api/conversations?page=1&status=resolved` with filters above.
| Conversation ID | Agent / Version | Channel | Application | Status | Outcome | Started | Duration | |
|---|---|---|---|---|---|---|---|---|
| C-4821 | Outbound Revenue Coach · v7 | Web | Sales | Resolved | Meeting booked · ACV $45K | 2025-10-06 08:21 | 12m | |
| C-4772 | Support Concierge · v5 | Support | Escalated | Billing dispute escalated to L2 | 2025-10-05 19:44 | 28m | ||
| C-4704 | Talent Screener · v3 | Voice | ATS | Resolved | Candidate disqualified · Notes logged | 2025-10-03 13:02 | 33m |
Transcripts Archive
Compliance-ready storage of every conversation. Mirrors `conversations`, `messages`, and `demo_conversations` tables with export controls.
Metadata Snapshot
Source: `conversations.metadata`Key Attributes
- Intent: Expansion · Sentiment: Positive · Confidence 0.92
- Tokens: 1,482 · Latency avg 340ms
- Artifacts: Proposal PDF, Quote #Q-193
Retention Policy
Copy transcripts to S3 with encryption at rest (KMS). Auto purge after 24 months unless flagged.
Backend Placeholder
Hook `GET /api/conversations/{id}` and `/api/conversations/{id}/messages` to hydrate export payloads and analytics.
Artifacts
Quote PDF
Generated from conversation C-4821. Stored in S3 with signed URL.
Scorecard JSON
ATS interview scoring from agent version v3. Linked to `demo_conversations` row.
Archived Conversations
Demo data from `demo_conversations`| Demo | Application | Highlights | KPIs | |
|---|---|---|---|---|
| High-Value Enterprise Lead | Sales | Budget $50K · Meeting booked · Objections overcome | Lead score 95 · Probability 85% | |
| Technical Issue Resolved | Support | Login fix for iOS · Sentiment from negative to neutral | FRT 2m · MTTR 8m |
Conversations
Pre-populated Demos
Interactive, KPI-aligned demos curated for quick walkthroughs across every application. Load transcripts, JSON scorecards, and KPI callouts without touching production data.
PHP Integration Placeholder
Wire up `GET /api/conversations/demos` & `GET /api/conversations/demos/{id}` to hydrate this UI with live data.
Sales Agent
6
BANT-qualified transcripts with ACV & forecast KPIs.
Customer Support
5
FRT, MTTR, and SLA coverage across multi-channel tickets.
AI Interviewer
4
ATS scorecards with stage conversion & candidate experience.
AI Mock Interviewer
4
Practice debriefs with rubric, engagement, and improvement KPIs.
Demo Collections
Demo Library
Hydrate from `GET /api/conversations/demos`Need something specific?
Generate custom demo scripts and JSON using Playground → Save to `demo_transcripts` via PHP `POST /api/conversations/demos`.
Selected Demo Overview
Showing High Value Lead — TechFlow Enterprises
Lead Source
web
Time-to-first-response
3 seconds
Opportunity Amount
$200,000
Forecasted Revenue
$140,000
KPI Snapshot
- Lead Status
- qualified
- SQL Reached
- true
- Opportunity Stage
- proposal
- Meeting Scheduled
- 2025-10-15T14:00:00-07:00
Conversation Summary
IT Director James Chen needs 50 high-performance workstations with a $200k budget and Q1 2025 timeline. Agent Alex confirms all BANT criteria, creates a proposal-stage opportunity, and books a follow-up demo.
Transcript Excerpt
Served via `GET /api/conversations/demos/{id}/transcript`Agent: Hello! This is Alex from DoBeyond Solutions. How can I help you today?
Customer: Hi Alex, I'm James Chen, IT Director at TechFlow Enterprises. We're looking to upgrade our office infrastructure with new workstations.
Agent: Great to meet you, James! I'd be happy to help. Can you tell me more about what you're looking for?
Customer: We need about 50 high-performance workstations for our development team...
JSON KPI Output
Sync scorecard via `POST /api/conversations/demos/{id}/scorecard`{
"lead_created": true,
"lead_status": "qualified",
"mql_reached": true,
"sql_reached": true,
"qualification_criteria": {
"budget_confirmed": true,
"authority_confirmed": true,
"need_confirmed": true,
"timeline_confirmed": true
},
"first_response_timestamp": "2025-10-12T09:15:03Z",
"time_to_first_response_seconds": 3,
"opportunity_created": true,
"opportunity_stage": "proposal",
"opportunity_amount": 200000,
"opportunity_probability": 70,
"opportunity_close_date": "2025-12-31",
"deal_closed": false,
"forecasted_revenue": 140000,
"meeting_scheduled": true,
"meeting_date": "2025-10-15T14:00:00-07:00",
"next_action": "Present detailed proposal at Thursday meeting",
"customer_name": "James Chen",
"customer_email": "[email protected]",
"customer_company": "TechFlow Enterprises",
"customer_title": "IT Director",
"objections": [],
"competitor_mentioned": "Dell",
"budget_range": "$200,000",
"timeline": "Q1 2025",
"outcome": "opportunity_created"
}
Upcoming Releases
- Spanish-language sales discovery with localized KPIs.
- Omni-channel support escalation featuring Zendesk + Salesforce sync.
- Executive panel interview demo with DEI & bias flag annotations.
- System design mock session with rubric auto-scoring.
Knowledge Sources
Documents & Files
Ingest files, URLs, and connected systems into the `knowledge_sources` table. Each source flows through chunking and embedding jobs before becoming searchable context for agents.
Active Sources
38
Files, URLs, CMS feeds stored in `knowledge_sources`.
Sync Health
92%
Successfully synced in the last 24h.
Embeddings
128K
Rows in `knowledge_items` with pgvector embeddings.
Source Manager
Backed by `POST /api/knowledge/sources`Processing Flow
- Persist metadata to PostgreSQL (`knowledge_sources`).
- Stream file to object storage for chunking.
- FastAPI worker calls `/rag/embed` to create vectors.
- Store chunks + embeddings in `knowledge_items` with ACL defaults.
Ingestion Queue
Monitor pending work pulled by background workers. Connect to PHP `GET /api/knowledge/jobs`.
Source: Playbook.pdf · Chunks: 142 · Duration: 48s
Updated `knowledge_items` at 2025-10-11 09:18.
Source: Partner FAQ URL · Awaiting download
Trigger worker via `POST /api/knowledge/jobs/{id}/run`.
Backend Placeholder
Expose job status from Redis/Queue for ingestion and embedding tasks.
Knowledge Sources Table
Synced from `knowledge_sources`| Source | Type | Sync Frequency | Last Synced | Status | |
|---|---|---|---|---|---|
|
Customer Onboarding Guide.pdf workspace_id: 9a2c-... |
File | Daily | 2025-10-11 08:12 | Active | |
|
Support Help Center type: cms · workspace_id: 2fd4-... |
CMS | Hourly | 2025-10-11 09:45 | Syncing | |
|
Renewal Playbook (Google Drive) sync_frequency: manual |
URL | Manual | 2025-10-09 14:02 | Action Required |
Grant document-level permissions via `knowledge_sources.access_control` after ingestion.
Chunk Preview
Inspect `knowledge_items` generated by the embedding pipeline.
| Chunk ID | Source | Preview | Roles |
|---|---|---|---|
| 194 | Onboarding Guide | "Kick-off objectives include aligning success metrics, provisioning SSO, and enabling the Revenue Workspace..." | Owner, Admin |
| 889 | Case Study | "Automation coverage increased 34% after deploying agent-led onboarding journeys..." | Owner, Sales-Manager |
Embeddings stored in `knowledge_items.embedding` (pgvector 1536-dim).
Source Integrations
Enrich the knowledge base with automated connectors.
-
Google Drive
Webhook polling · Syncs shared folders.
-
Notion Knowledge Base
Pulls published pages via API. Maps tags to metadata.
-
Zendesk Help Center
Sync frequency hourly · Mirrors articles table.
-
Confluence Spaces
Applies chunking recipe: 800 tokens w/ 20% overlap.
API Placeholder
Persist connector credentials via PHP `POST /api/knowledge/integrations` and schedule sync jobs.
RAG Collections
Bundle sources for targeted retrieval. Mirrors the `rag_collections` table and populates agents with curated context.
| Collection | Linked Agent | Sources | Config | Last Updated | |
|---|---|---|---|---|---|
| Sales Master Collection | Outbound Revenue Coach | Playbook.pdf, Case Studies, CRM Opportunities | top_k 5 · threshold 0.72 | 2025-10-08 | |
| Support Knowledge Core | Support Concierge | Help Center CMS, Product Docs | top_k 8 · threshold 0.68 | 2025-10-05 |
Semantic Search Test
Run ad-hoc queries through the FastAPI `/rag/search` endpoint.
Top Results
- Chunk #194 · Playbook.pdf · Confidence 0.84
- Chunk #889 · Case Study · Confidence 0.79
Access Control
Configure visibility per role via `knowledge_items.access_control`.
Backend Placeholder
Hook to PHP `PUT /api/knowledge/sources/{id}` to persist ACL changes.
Articles & FAQs
Publish support content with AI-assisted drafting. Leverages knowledge sources and surfaces analytics.
Article Drafts
AI Copilot Adoption FAQ
Guides new customers on enabling copilots. Drafted via prompt templates.
Escalation Runbook
Step-by-step workflow for billing escalations with embedded links.
Performance
Deflection Rate
48%
+6% vs last week
Top Article
AI Copilot Adoption FAQ
2.1K views
Draft Completion
12
Queued for review
Backend Placeholder
Use PHP endpoints to fetch analytics and update article status. Connect to the RAG service for AI-generated drafts.
Model Readiness
Training Data
Assemble labeled corpora for fine-tuning and guardrail evaluation. Datasets track provenance from `knowledge_items`, capture labels, and feed batch jobs across your PHP + FastAPI pipeline.
Datasets
12
Active datasets ready for fine-tuning.
Label Coverage
78%
Items with reviewer-approved tags.
Guardrail Tests
5
Scheduled evaluations awaiting review.
Dataset Builder
Persists via `POST /api/training/datasets`Sampling Strategy
- Filter chunks by metadata tag `topic='renewal'`.
- Balance positive / negative sentiment examples (50/50).
- Exclude items flagged `access_control.roles` without reviewer permission.
Training Jobs
Sync job status from PHP `GET /api/training/jobs`.
Dataset: Sales Objection Handling · F1: 0.87 · Toxicity Violations: 0
Completed 2025-10-10 21:15 via FastAPI `/eval/run`.
Dataset: Support FAQ Ground Truth · Model: gpt-4o-mini
Streaming logs from PHP queue. ETA 18m.
Backend Placeholder
Use PHP `POST /api/training/jobs` to enqueue runs and poll FastAPI for status updates.
Labeling Queue
Stores reviewer feedback| Item | Prompt | Desired Output | Labeler | |
|---|---|---|---|---|
| Item #582 | "Respond to budget objection..." | Use ROI framing with two bullet points. | Alex Rivera | |
| Item #593 | "Clarify invoice discrepancy..." | Flag for escalation if customer rating < 3. | Taylor Reed |
Persist reviewer decisions via `PUT /api/training/items/{id}`.
Compliance & Audit
PII Scan
Latest sweep detected 0 redacted fields. Run PHP `POST /api/training/audit/pii` nightly.
Bias Review
Ensure protected attributes are excluded. Store findings in `training_audit_logs`.
API Placeholder
Expose audit trails via `GET /api/training/audit?dataset_id=` for exports.
Visual Workflow Builder
Design event-driven automations from triggers to actions. Backed by the `workflows` table and publish endpoints like `POST /api/workflows`.
Workflow Outline
Draft (unsaved)API Hook
Persist to PHP backend using `POST /api/workflows`. Conditions and actions serialize into JSONB columns (`conditions`, `actions`).
Builder Canvas
Trigger: Conversation Ended
Scope: Sales application · Condition: sentiment != negative
Condition
`metadata.outcome` = "Meeting Booked" OR `token_count` < 1500
Action 1: notify_slack
Channel #sales-wins · Template `renewal_playlist`
Action 2: schedule_meeting
Use `calendar_booking` tool · Owner: Taylor Reed
Fallback
If any action fails, escalate via `create_ticket` with severity high.
Triggers & Rules
Define conditional logic that powers automations. Stores expressions inside `workflows.conditions` JSONB.
Rule Library
High Intent Lead
Trigger: `lead_created` · Condition: `score > 80` AND `source = "demo"`
SLA Breach
Trigger: `ticket_priority_high` · Condition: `waiting_minutes > 30`
Logic Builder
If
`conversation.sentiment` = "negative"
And
`workflow.trigger_data.customer_tier` = "enterprise"
Then
Route to escalation queue · `escalate_to_queue`
Backend Placeholder
Serialize rule into JSON and persist via `PUT /api/workflows/{id}`.
Actions Library
Reusable automation blocks with audit history, versioning, and API references.
| Action | Type | Status | Last Sync | |
|---|---|---|---|---|
| Send Welcome Email | Active | 2025-10-07 | ||
| Notify Slack Channel | Slack Notification | Active | 2025-10-08 | |
| Create Support Ticket | Ticket Creation | Active | 2025-10-09 |
Action Details
Send Welcome Email
Sends a welcome email to new leads. Template stored in `email_templates` table.
Linked Agents: Outbound Revenue Coach, Support Concierge
Notify Slack Channel
Sends a notification to a specific Slack channel. Template stored in `slack_templates` table.
Linked Agents: Outbound Revenue Coach, Support Concierge
Create Support Ticket
Creates a support ticket in Zendesk. Fields populated from conversation data.
Linked Agents: Support Concierge
Workflow History
Trace automation runs, debug agent decisions, and export execution timelines.
Enterprise Pulse
Executive Dashboard
Cross-application snapshot for CRM, Support, ATS, and Learning apps. Data hydrates via PHP aggregations such as GET /api/analytics/executive?scope=quarter.
$6.8M
Pipeline influenced by Sales Agent automations.
`GET /api/analytics/crm/pipeline-summary`
14 min FRT
Customer Support app maintains MTTR 3.4h.
`GET /api/analytics/support/sla-overview`
26d
Average time-to-hire powered by AI Interviewer.
`GET /api/analytics/ats/time-to-hire`
74%
Average rubric uplift in AI Mock Interviewer cohorts.
`GET /api/analytics/mock/improvement-trend`
Application Health Overview
Synced nightly via `/api/analytics/app-health`Alerts & Actions
-
Support queue approaching SLA breach
Trigger workflow via `POST /api/alerts/acknowledge`
-
ATS diversity funnel dipped 3pts WoW
Check sourcing split —
GET /api/analytics/ats/diversity -
Mock interviews backlog flagged
Assign new sessions via coaching scheduler.
Backend Placeholder
Use PHP job to persist alert acknowledgements in `executive_alerts`.
Revenue & Forecast
Lead acceptance rate 37% vs baseline 24%.
Experience Scores
Updated hourlyPowered by Customer Support surveys.
AI Interviewer follow-up surveys.
Data Refresh Placeholder
Bind to PHP cache layer exposing `GET /api/analytics/experience`.
Operational Insights
Application Metrics
Dive deeper into KPIs for each DoBeyond application. Panels hydrate via PHP routes such as GET /api/analytics/applications?app=sales.
Lead → MQL
46%
AI intake + enrichment uplift +9pts vs last month.
SQL Win Rate
28%
Alignment with playbooks; forecast accuracy 91%.
Time to First Response
1.6h
Slack + WhatsApp automations reduce delay by 34%.
Average Contract Value
$38K
Quote AI ensures discount discipline.
Pipeline Quality
`GET /api/analytics/crm/pipeline-quality`Top-performing industries: SaaS, Logistics.
Placeholder
Bind to PHP endpoint caching weighted forecasts.
Recent Deals
| Opportunity | Stage | Amount |
|---|---|---|
| Acme Renewals | Negotiation | $420K |
| Globex Expansion | Proposal | $180K |
| BlueSky Pilot | Commit | $95K |
Rendered via `/api/analytics/crm/deals?limit=5`
Deflection Rate
41%
LLM front-door resolves 4.8k tickets monthly.
First Response Time
14m
SLA target < 20m; 93% attainment.
Mean Time to Resolve
3.4h
Automation triage reduces escalations -11%.
CSAT
4.4 / 5
Macro generator keeps tone on-brand.
Queue Load
`GET /api/analytics/support/queues`| Queue | Backlog | SLA | Breaches |
|---|---|---|---|
| Tier 1 Chat | 128 | 95% | 3 |
| Tier 2 Technical | 62 | 88% | 7 |
| Priority Incidents | 12 | 100% | 0 |
Knowledge Impact
-
Articles Updated
34 revisions via
POST /api/support/articles/sync -
Macro Adoption
78% of replies templated with AI suggestions.
-
Backend Placeholder
Expose helpfulness trend via `/api/analytics/support/articles`.
Time to Hire
26 days
Down 4 days WoW thanks to async interviews.
Stage Conversion
64%
Qualified candidates hitting onsite stage.
Candidate Experience
+48 NPS
Bias detection surfaces fairness notes.
Offer Acceptance
81%
Offer AI ensures comp alignment.
Pipeline Health
`GET /api/analytics/ats/pipeline`Source effectiveness: Referrals 32%, Job Boards 27%, AI Sourcing 21%.
Placeholder
Hook into `/api/analytics/ats/diversity` for cohort tracking.
Upcoming Interviews
| Candidate | Stage | Scheduled |
|---|---|---|
| Sara Chen | Panel Interview | Oct 15 · 11:00 |
| Luis Gomez | Behavioral | Oct 16 · 09:30 |
| Aisha Khan | Offer Review | Oct 16 · 16:00 |
Feed via `/api/analytics/ats/interviews`.
Session Completion
88%
Students completing ≥3 sessions per week.
Rubric Improvement
+18 pts
Communication & structure scores week-over-week.
Engagement Minutes
6.2K
Conversation pace + filler word tracking.
Placements
39
Opt-in student successes shared with employers.
Skill Radar
`GET /api/analytics/mock/skills`STAR responses +16%, technical depth +12%.
Placeholder
Connect to `/api/analytics/mock/cohorts` for instructor filters.
Assignments & Badges
-
Active Assignments
62 pending — `GET /api/mock/assignments?status=pending`
-
Badges Issued
128 new badges this month.
-
Backend Placeholder
Sync badge evidence via `/api/mock/badges`.
Evaluation Hub
Model Performance
Monitor accuracy, latency, safety, and drift across the AI agents powering CRM, Support, ATS, and Mock Interview workflows.
Prediction Accuracy
94%
Sales routing agent vs human verified outcomes.
`GET /api/models/metrics?model=sales-router`
Support Resolution Quality
4.6 / 5
LLM suggested macros audited by QA analysts.
`GET /api/models/metrics?model=support-copilot`
Bias Watch
+0.7σ
ATS interviewer fairness delta vs threshold ±2σ.
`GET /api/models/metrics?model=ats-interviewer`
Latency (p95)
2.1s
Mock interviewer multimodal responses.
`GET /api/models/metrics?model=mock-coach`
Evaluation Benchmarks
`GET /api/models/evals?benchmark=production`| Model | Task | Score | Trend |
|---|---|---|---|
| Sales Routing v3 | Intent + slot extraction | 0.94 F1 | ▲ |
| Support Concierge 2.1 | Resolution suggestions | 4.6 QA | ▲ |
| ATS Interviewer | Async interview scoring | 0.88 κ | ▶ |
| Mock Coach | Feedback generation | 4.7 rating | ▲ |
Quality Notes
-
CRM intents exceeding 0.9 F1
90th percentile intents: qualification, pricing, objections.
-
Support summarization refresh
New prompt reduces hallucinations by 38%.
-
Backend Placeholder
Use `/api/models/evals/notes` for PHP-managed review queue.
Drift Signals
`GET /api/models/drift?group=applications`Spike in billing issues — suggest new macro training.
Feature drift flagged on budget signals — monitored.
Placeholder
Connect to `/api/models/drift/alerts` for automated mitigations.
Guardrail Triggers
| Timestamp | Guardrail | Action |
|---|---|---|
| Oct 12 · 08:14 | Toxicity > 0.7 | Escalated to human agent |
| Oct 11 · 19:42 | PII Detected | Redacted + stored in vault |
| Oct 11 · 09:17 | Policy Deviation | Auto re-generated response |
Stream via `/api/models/guardrails/logs`.
Latency & Cost
`GET /api/models/latency?window=24h`Cache hit rate 71% via `/api/cache/status`.
Streaming responses enables < 2s for 82% of chats.
Placeholder
Attach cost-per-token data via `/api/models/costs`.
Scaling Controls
-
Autoscaling windows
`POST /api/models/scale` to adjust concurrency.
-
Cache refresh cadence
Preload prompts hourly for CRM + Support.
-
Backend Placeholder
Model cost budgets from `/api/models/billing`.
Offline Experiments
`GET /api/models/experiments`| Experiment | Variant | Metric | Delta |
|---|---|---|---|
| CRM Prompt v3 | Few-shot | Lead accuracy | +3.2% |
| Support Summaries | Guarded | Hallucinations | -38% |
| ATS Bias Mitigation | Re-weighted | Fairness gap | -1.1σ |
Live A/B Tests
-
ATS Follow-up Prompt
Variant B booking rate +7% (p=0.03).
-
Mock feedback tone
Soft coaching style increases session repeats.
-
Backend Placeholder
Write experiment outcomes to `/api/models/experiments/{id}`.
Custom Reports
Build tailored dashboards with exports, scheduled delivery, and blended datasets.
Integration Hub
Connected Apps
Centralize CRM, support, hiring, communication, and storage connectors. Data is stored in the `integrations` table and orchestrated through PHP routes such as `GET /api/integrations`, `POST /api/integrations/{id}/sync`, and `DELETE /api/integrations/{id}`.
Active Integrations
18
Across CRM, CS, ATS, and comms stacks.
Last Sync
12m
HubSpot opportunities refresh.
Errors
2
Zendesk webhook retries queued.
Upcoming
6
Scheduled sync windows within the hour.
Integration Directory
Manage via `GET /api/integrations?type=`| Connector | Category | Sync Mode | Last Sync | Health | |
|---|---|---|---|---|---|
| Salesforce | CRM | Bi-directional | 2025-10-12 08:12 | Healthy | |
| Zendesk | Support | Two-way tickets | 2025-10-12 07:58 | Warning | |
| Greenhouse | ATS | Daily import | 2025-10-11 23:45 | Healthy | |
| Twilio | Voice/SMS | Event-driven | 2025-10-12 08:00 | Monitoring |
Use `POST /api/integrations/{id}/trigger` to run ad-hoc syncs.
Calendars & Email
Calendars + Gmail threading with Conversation IDs.
Last Free/Busy pull 3 minutes ago.
Outlook + Teams scheduling. Renew OAuth token via `POST /api/integrations/{id}/refresh`.
Storage & Artifacts
Manage S3, GCS, and Azure Blob connectors for transcript and asset storage.
Amazon S3
Bucket: dobeyond-artifacts · Region: us-east-1
Azure Blob
Used for EU data residency. Signed URLs expire in 15m.
Backend Placeholder
Expose storage health via `GET /api/integrations/storage/status`.
Event Delivery
API Webhooks
Configure outbound callbacks for CRM updates, transcript events, and automation triggers. Routes use PHP endpoints like `POST /api/webhooks`, `PUT /api/webhooks/{id}`, and `GET /api/webhooks/deliveries`.
Recent Deliveries
Events pulled from `webhook_deliveries`.
| Time | Event | Status | Latency |
|---|---|---|---|
| 08:12:44 | conversation.completed | 200 OK | 620ms |
| 08:05:11 | lead.created | Retrying | -- |
| 07:58:02 | ticket.updated | 200 OK | 410ms |
Fetch full history via `GET /api/webhooks/deliveries?webhook_id=`.
Header Templates
Store signature headers for partner verification.
HubSpot Listener
Header: `X-HubSpot-Signature` · Hash: SHA256
Custom Partner
Header: `X-DoBeyond-Signature` · Hash: HMAC-SHA1
Delivery Rules
Guard against runaway retries.
- Max concurrency 5
- Dead-letter queue Enabled
-
Backend Placeholder
Configure via `PUT /api/webhooks/policies`.
Channel Orchestration
Channel Configurations
Enable voice, chat, email, and messaging across Twilio/Infobip, WhatsApp, SIP/SIPREC, and web chat. Settings persist in `channel_configs` and PHP endpoints such as `POST /api/channels`, `PUT /api/channels/{id}`, and `GET /api/channels/status`.
Channel Matrix
`GET /api/channels`| Channel | Provider | Workspace | Status | |
|---|---|---|---|---|
| Voice (PSTN) | Twilio SIPREC | Global HQ | Recording On | |
| Messaging | WhatsApp via Infobip | Support | Limited Beta | |
| Gmail API | Sales APAC | Threading Enabled | ||
| Calendar | Google / Microsoft | All | Free/Busy |
Update through `PUT /api/channels/{id}`. PSTN recordings stored in S3 with signed URLs.
Voice Routing
Configure SIP trunking and WebRTC handoff.
Inbound PSTN
Numbers: +1 (628) 555-1024 (US), +44 20 3322 5544 (EU)
Routing: AI front-line → escalation queue via `POST /api/channels/voice/routes`.
Call Recording
SIPREC streams saved to S3 with TLS. Transcripts piped into knowledge ingestion.
Backend Placeholder
Provision voice channels via `POST /api/channels/voice` and confirm SIP credentials asynchronously.
Messaging Templates
Sync approved WhatsApp templates and SMS snippets.
Renewal Reminder
Language: en-US · Status: Approved
Support Follow-up
Language: es-MX · Status: Pending
Channel Guardrails
Define escalations, fallback intents, and quiet hours.
- Quiet Hours 22:00–06:00
- Escalation Threshold Sentiment < -0.6
-
Backend Placeholder
Persist guardrails via `PUT /api/channels/guardrails`.
Partner Ecosystem
Integration Marketplace
Discover certified connectors with guided provisioning. Marketplace listings live in `integration_catalog` and surface install flows via PHP `POST /api/integrations/install`.
HubSpot
Pipeline sync + enrichment
Two-way sync for contacts, deals, and activities. Auto map conversation summaries to HubSpot notes.
Zendesk
Support ticket orchestration
Sync transcripts to tickets, push AI summaries, and respect Zendesk macros.
Greenhouse
ATS candidate sync
Auto-create scorecards from AI interviews and push status updates back to hiring pipelines.
Outlook & Teams
Email + calendar + meetings
Bi-directional calendar invites, thread replies with conversation IDs, and push call recordings to Teams.
Workday Export
Enterprise HRIS bridge
Batch export interview scores and onboarding data to Workday staging tables.
Submit Integration Idea
Have a connector in mind? Send specs to our partner team.
Organization Settings
Configure tenant-level metadata, internal tiers, and platform defaults. Backed by the `organizations` table with PHP API endpoints such as `POST /api/auth/register` and `GET /api/organizations/current`.
API Endpoints
PHP (Laravel) gateway-
POST /api/auth/registerProvision -
GET /api/organizations/currentRead -
PUT /api/organizations/{id}Update
Backend Placeholder
Connect this form to your PHP service. On submit, call the relevant API using `fetch('/api/organizations/{id}', { method: 'PUT', body: JSON.stringify(payload) })` and surface success/error toast messages.
Organization Metadata
Reflects `organizations` table columns for quick auditing.
| Field | Sample Value | Description |
|---|---|---|
| id | f1d6-8c42-48a3-9211 | Primary key UUID generated via `gen_random_uuid()` |
| name | DoBeyond Labs | Display name for tenant and invoices |
| slug | dobeyond | Unique slug used for subdomain routing |
| tier | growth | Billing / entitlement tier (starter, growth, enterprise) |
| settings | {"branding": "dark", "ai_guardrails": true} | JSONB configuration for feature flags and preferences |
| created_at | 2025-01-08 14:14:22 | Timestamp of tenant creation |
Users & Roles
Manage members across workspaces with hierarchical RBAC. Powered by the `users` table and endpoints like `POST /api/users` (admin) and `PUT /api/users/{id}/role`.
| User | Role | Workspaces | Last Login | Status | ||
|---|---|---|---|---|---|---|
| Taylor Reed | [email protected] | Owner | Global HQ | 2025-10-10 08:23 | Active | |
| Priya Chen | [email protected] | Admin | Global HQ, Sales APAC | 2025-10-09 19:02 | Active | |
| Marco Alvarez | [email protected] | Manager | Support Americas | 2025-10-05 12:17 | Pending |
Role Permissions Matrix
Maps to `users.role` and policy checks| Capability | Owner | Admin | Manager | Agent | Analyst |
|---|---|---|---|---|---|
| Manage organization | — | — | — | ||
| Invite users | — | — | |||
| Configure automations | — | ||||
| View analytics |
Workspaces
Segment teams within an organization for data isolation. Mirrors the `workspaces` table with `GET /api/workspaces` for listing and `POST /api/workspaces` for creation.
| Workspace | Organization | Created | Members | Status | |
|---|---|---|---|---|---|
| Global HQ | DoBeyond Labs | 2025-02-18 | 42 | Active | |
| Sales APAC | DoBeyond Labs | 2025-04-02 | 18 | Active | |
| Support Americas | DoBeyond Labs | 2025-03-12 | 27 | At Risk |
Identity & SSO
Authentication Settings
Manage registration flows, SSO providers, and tenant-level login policies. Backs the `organizations`, `workspaces`, and `users` tables plus PHP endpoints like `POST /api/auth/register`, `POST /api/auth/sso/initiate`, and `GET /api/organizations/current`.
Login Policies
Persist via `PUT /api/auth/policies`Passwordless Login
Enable magic links and passkeys for all users.
Require MFA
Force TOTP or hardware keys for admins and owners.
Audit Trail Preview
Recent events: Taylor enabled MFA (2025-10-11 09:18), Priya added Okta SSO (2025-10-10 14:05).
SSO Providers
Configure OAuth / SAML connections. Use `POST /api/auth/sso/initiate` for metadata exchange.
Client ID: okta-7821 · Domains: dobeyond.ai
Client ID: azure-9912 · Metadata upload required
Backend Placeholder
Expose SSO status via `GET /api/auth/sso/providers` and allow admins to revoke credentials.
Provisioning
Just-in-time (JIT) Users
Automatically create user records on first SSO login. Map SSO groups to internal roles using `users.role`.
SCIM Provisioning
Support automatic user lifecycle management via `/api/auth/scim/*` endpoints.
Activity Log
Monitor identity events stored in `auth_audit_logs`.
| Timestamp | User | Event | Details |
|---|---|---|---|
| 2025-10-11 09:18 | Taylor Reed | MFA Enabled | Enforced TOTP for Owner role. |
| 2025-10-10 14:05 | Priya Chen | SSO Added | Configured Okta integration. |
| 2025-10-09 21:42 | Marco Alvarez | Login Blocked | Exceeded retry attempts (5). |
Expose full logs via `GET /api/auth/audit?limit=100`.
Security & Compliance
Trust & Governance Center
Monitor posture across SOC 2, ISO 27001, HIPAA, and GDPR controls. Surfaces data from `security_audits`, `retention_policies`, and `access_logs` with PHP endpoints such as `GET /api/security/posture`, `POST /api/security/audits`, and `GET /api/security/events`.
SOC 2
Compliant
Policies synced 2025-09-30.
ISO 27001
In Review
Control A.12.6 pending evidence.
Incidents
0
Last 30 days
DLP Scans
342
Redactions applied automatically.
Access Control Policies
`PUT /api/security/acl`Define per-object permissions across roles. Fine-grained ACL updates sync to the `acl_policies` table and propagate to downstream services.
| Object | Owner | Admin | Manager | Analyst | Agent |
|---|---|---|---|---|---|
| Agents | — | — | |||
| Knowledge Base | — | ||||
| Transcripts | — | ||||
| Exported Reports | — | — |
Backend Placeholder
Wire this matrix to a PHP controller (e.g., `SecurityAclController@update`). POST payload mirrors the displayed rows and actions.
Field-Level Redaction
Apply masking and hashing to sensitive fields before storage or export. Redaction profiles map to agent context windows.
| Field | Rule | Preview |
|---|---|---|
| salary_expectation | Dynamic mask (retain currency) | $***,*** |
| candidate_email | Hash w/ salt | 8c52a9•• |
| credit_card_last4 | Redact unless PCI scope | XXXX |
Backend Placeholder
Expose rule definitions through `GET /api/security/redaction` and persist edits via `PATCH /api/security/redaction/{field}`.
Region & Retention Matrix
`PUT /api/security/residency`Map every dataset to a primary and failover region with retention windows per object type.
| Dataset | Primary Region | Failover | Retention |
|---|---|---|---|
| Conversations | eu-central-1 | us-east-1 | 24 months |
| Knowledge Chunks | ap-southeast-1 | ap-southeast-2 | 12 months |
| Audit Logs | us-gov-west-1 | eu-west-2 | 36 months |
Backend Placeholder
Surface residency options from `region_profiles` and drive updates through a Laravel job queue for data movement.
Prompt Archives & PII Masking
`POST /api/security/archives`Archive prompts and responses for audit-ready history while masking customer PII.
Backend Placeholder
Connect to PHP worker that streams prompt archives into object storage. Provide toggle state via `GET /api/security/archives/settings`.
Human Oversight Controls
`PUT /api/security/hitl`Set thresholds that determine when an agent handoff is required before customer delivery.
Messages below this score move to manual review.
Pricing exceptions, legal claims, refund over $5K.
Backend Placeholder
Pass slider values to PHP services that compute pending approvals. Store overrides in `agent_review_policies`.
Approval Queue Snapshot
Track pending human reviews before responses are dispatched.
| Item | Submitted | Reviewer | Status |
|---|---|---|---|
| Offer letter draft | 2025-10-11 20:17 | Taylor Reed | Awaiting |
| Refund escalation | 2025-10-11 18:02 | Priya Chen | Approved |
| Employment verification | 2025-10-11 09:41 | Marco Alvarez | Rejected |
Backend Placeholder
When ready, hydrate this table with `GET /api/security/review-queue?status=open` and provide action buttons for approve/reject.
Model Governance Registry
`GET /api/models`Track deployed models, version lineage, and promotion readiness before publish.
| Model | Version | Stage | Owner | Last Eval |
|---|---|---|---|---|
| candidate_matcher | v3.2.1 | Ready | Priya Chen | 2025-10-09 |
| offer_generator | v2.8.0 | Needs Regression | Taylor Reed | 2025-10-08 |
| routing_classifier | v1.5.4 | Active | Marco Alvarez | 2025-10-06 |
Backend Placeholder
Populate from `model_registry` table. Enforce governance with PHP service `POST /api/models/{id}/promote`.
Evaluation & Regression Suite
Manage evaluation datasets and regression runs before shipping new agent versions.
`datasets/hiring-core.jsonl` · Last updated 2025-10-07 by Priya Chen.
Run ID `reg-20251010` · `offer_generator` vs baseline. Drift < 2%.
Backend Placeholder
Hook up PHP endpoints `POST /api/models/{id}/regress` and `GET /api/models/evaluations` to automate dataset uploads and compare runs.
Control Checklist
Sync via `GET /api/security/controls`| Framework | Control | Status | Owner |
|---|---|---|---|
| SOC 2 | CC6.6 - Change Management | Pass | Priya Chen |
| ISO 27001 | A.12.6 - Technical Vulnerability | Pending | Taylor Reed |
| HIPAA | 164.308(a)(1) - Risk Management | In Progress | Marco Alvarez |
Upload evidence packages via `POST /api/security/audits/{id}/evidence`.
Threat Detection Feed
Stream anomalies from SIEM integrations.
Workspace: Support Americas · Source IP: 104.82.11.5
Escalate via `POST /api/security/incidents`.
Blocked by DLP policy · User: [email protected]
Review via `GET /api/security/events/{id}`.
Backend Placeholder
Integrate SIEM (e.g., Panther, Splunk) webhooks into PHP `POST /api/security/events` endpoint.
Data Retention
Align global deletion windows across tables.
| Dataset | Retention | Policy |
|---|---|---|
| Conversations | 24 months | Purged unless flagged |
| Knowledge Chunks | 12 months | Rotated on re-ingest |
| Audit Logs | 36 months | Immutable storage |
Update via `PUT /api/security/policies/retention`.
Compliance Requests
Automate DSAR, DPA, and audit documentation.
Responded 2025-10-09. Records exported from `users` and `conversations`.
Upload signed contract via `POST /api/security/requests/{id}/documents`.
Billing & Usage
Subscription Management
Track plan entitlements, usage-based spend, and invoices. Mirrors the `subscriptions`, `usage_metrics`, and `invoices` tables. Integrate with PHP endpoints like `GET /api/billing/summary`, `POST /api/billing/payment-method`, and `GET /api/billing/invoices`.
Current Plan
Growth
Seats included: 50 · Workspaces: 10
Usage This Month
$4,230
AI tokens + automation runs
Next Invoice
Nov 01
Estimated total $5,120
Usage Breakdown
`GET /api/billing/usage?range=30d`| Metric | Quantity | Unit Cost | Total |
|---|---|---|---|
| LLM Tokens | 32.4M | $0.00012 | $3,888 |
| Automation Runs | 840 | $0.25 | $210 |
| Premium Seats | 12 | $18 | $216 |
Schedule exports via `POST /api/billing/reports`.
Plan Actions
Request Upgrade
Move to Enterprise for unlimited workspaces and dedicated support.
Usage Alerts
Email and Slack alerts at 80% and 100% of quota.
API Placeholder
Attach Stripe customer portal via PHP `GET /api/billing/portal`.
Invoices
| Invoice | Period | Amount | Status | |
|---|---|---|---|---|
| INV-2025-09 | Sep 01 - Sep 30 | $4,812 | Paid | |
| INV-2025-08 | Aug 01 - Aug 31 | $4,596 | Paid | |
| INV-2025-07 | Jul 01 - Jul 31 | $4,402 | Processing |
Pull invoice PDFs from PHP `GET /api/billing/invoices/{id}`.
Consumption Forecast
Predict next month's spend using trailing averages.
Projected Total: $5,420
- Tokens: +12% vs last month
- Automations: steady
- Premium Seats: +2 upgrades pending
Forecast generated via `GET /api/billing/forecast?months=1`.