Skip to content

About

Agentic Governance Operating System — Visual AI-powered governance, compliance, security, and risk management platform

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

12 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sen'Analytics Mission Control

Agentic Governance Operating System

Sen'Analytics Mission Control is a production-grade visual operating system for governance, compliance, security, risk management, and data engineering operations — built for Senegal and Francophone West Africa.

It combines multi-agent orchestration, visual workflow building (n8n-style), real-time execution monitoring (Datadog-style), and comprehensive auditability (Purview-style) into a single platform.


Table of Contents


What It Does

Capability Description
Agent Orchestration 7 specialized AI agents (Supervisor → Discovery → Classification → Security → Compliance → Risk → Reporting) working through a supervisor-worker pipeline with human-in-the-loop approval checkpoints
Visual Workflows Drag-and-drop React Flow canvas with 6 custom node types (Agent, Tool, Approval, Policy, Condition, Trigger). Build governance pipelines visually
Real-Time Monitoring WebSocket-driven live execution streaming — nodes pulse green during execution, edges animate, metrics update in real-time
MCP Integrations Pluggable tool adapters for Snowflake, PostgreSQL, AWS, Azure, Databricks, Jira, ServiceNow, Power BI, Microsoft Purview, and SQL Server — 53 discoverable tools across 10 systems
Knowledge Graph Neo4j-powered governance graph — 29 nodes, 31 relationships, full lineage tracing, impact analysis, compliance mapping
Compliance Reporting Auto-generated executive summaries, audit reports, and remediation plans. Downloadable as TXT/JSON
Multi-Tenant RBAC Organization isolation with 4 roles (Admin, Editor, Viewer, Auditor), granular permissions
Full Auditability Immutable append-only audit logs, replayable workflows, traceable agent decisions

Quick Start

Prerequisites

  • Docker + Docker Compose v2+
  • Python 3.12+ + venv
  • Node.js 22+ + pnpm
  • Git

60-Second Start

# 1. Clone
git clone https://github.com/ibfaye/senanalytics-mission-control.git
cd senanalytics-mission-control

# 2. Infrastructure (once)
cd docker && docker compose up -d

# 3. Backend
cd ../backend
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

# 4. Frontend (new terminal)
cd ../frontend && pnpm install && pnpm dev --turbopack -p 3000

# 5. Open
open http://localhost:3000/mission-control

URLs

Service URL Credentials
Mission Control http://localhost:3000 —
API Docs (Swagger) http://localhost:8001/docs —
pgAdmin http://localhost:5050 admin@senanalytics.com / admin
Neo4j Browser http://localhost:7474 neo4j / senanalytics_dev

Architecture

System Layers

┌─────────────────────────────────────────────┐
│ L6 · Next.js 16 + React Flow Canvas          │ Frontend (15 pages)
├─────────────────────────────────────────────┤
│ L5 · FastAPI + WebSocket Server              │ API Gateway (47 endpoints)
├─────────────────────────────────────────────┤
│ L4 · Agent Mesh (7 agents, direct calls)     │ Orchestration
├─────────────────────────────────────────────┤
│ L3 · MCP Tool Layer (10 adapters)            │ External Systems
├─────────────────────────────────────────────┤
│ L2 · Neo4j · PostgreSQL · Redis              │ Data Layer
├─────────────────────────────────────────────┤
│ L1 · Docker · Kubernetes                     │ Infrastructure
└─────────────────────────────────────────────┘

Data Flow

User → Canvas → REST API → Engine → Agent Mesh → MCP Tools → External Systems
                    ↑                        ↓
                    └── WebSocket ←── Live Status (node.started/completed/failed)

Screen-by-Screen

1. Mission Control (/mission-control)

The main dashboard. Shows active workflows, running agents, live stats (executions, agents, costs, risks). Stats grid, recent workflow list with status badges, execute button per workflow.

2. Workflows (/workflows)

List all governance workflows with create, edit, and delete. Click any workflow to open the canvas.

3. Workflow Canvas (/workflows/[id])

The central experience — a React Flow canvas with:

  • Toolbar: Drag-and-drop node palette (Agent, Tool, Approval, Policy, Condition, Trigger) + Save/Execute buttons
  • Canvas: Infinite pan/zoom, minimap, background grid
  • Live Execution: Nodes animate during execution — pulsing green border → solid green with metrics
  • Approval Dialog: Modal overlay when execution hits a human checkpoint
  • Progress Bar: Shows "Executing 2/4 agents" during runs

4. Agents (/agents)

Registry of all 7 agents with type badges, descriptions, and status indicators.

5. MCP Servers (/mcp)

Connect external systems. Shows all 10 supported adapters with connection status, tool discovery (53 tools), connect/disconnect buttons, and a system grid with status indicators.

6. Knowledge Graph (/knowledge)

Graph explorer showing:

  • Node type distribution (Domains, Tables, Columns, Policies, Controls, Risks)
  • Risk heatmap with likelihood×impact score bars
  • Lineage trace: Customer Data → Customer 360 → customers → email → CDP-ART-34 → Control-Enc-01
  • Compliance gaps: PII columns with no mapped control

7. Governance Dashboard (/governance)

Data domains, products, classification stats, PII encryption status, lineage depth, data stewardship.

8. Security Dashboard (/security)

Security findings (encryption, IAM, network), risk heatmap with severity badges, compliance gap counts, control implementation status.

9. Compliance Dashboard (/compliance)

5-framework score cards (GDPR, CDP-Senegal, PIPEDA, SOC2, ISO27001) with status, gap counts, policy/control mappings, and a detail table.

10. Reports (/reports)

Generate 3 report types: Executive Summary, Audit Report, Remediation Plan. Download as TXT or JSON. Reports include findings, risk assessments, and prioritized action items.

11. Audit Trail (/audit)

Immutable log of all actions — workflow executions, agent decisions, approval events.

12. Organization (/organization)

Multi-tenant management — view org, members with role badges (Admin/Editor/Viewer/Auditor), RBAC permission reference grid.

13. Profile (/profile)

User profile with role, organization context, and API header display for multi-tenant scoping.


Agent Mesh

Supervisor-Worker Pipeline

┌──────────────┐
│  Supervisor  │  Plans → Routes → Delegates → Aggregates
└──────┬───────┘
       │
  ┌────┴────────────────────────────┐
  │                                  │
  ▼                                  ▼
┌──────────┐                   ┌──────────────┐
│ Discovery │                   │Classification│
│ 3 sources │                   │ 12 PII fields │
└────┬─────┘                   └──────┬───────┘
     │                                │
     └────────┬───────────────────────┘
              ▼
        ┌──────────┐
        │ 🛑 Human  │  ← Approval checkpoint
        │ Approval  │
        └────┬─────┘
             │
        ┌────┴────┐
        ▼         ▼
   ┌─────────┐ ┌──────────┐
   │Compliance│ │Reporting │
   │5 frameworks│ │3 reports │
   └─────────┘ └──────────┘

Verified Run

✅ Discovery:       450ms, 320 tok, $0.01
✅ Classification:  680ms, 540 tok, $0.02
🛑 Human Approval: Paused → Resolved via API
✅ Compliance:      750ms, 610 tok, $0.02
✅ Reporting:       380ms, 290 tok, $0.01
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📊 1,760 tokens · $0.06 · 2,260ms

MCP Tool Layer

10 System Adapters — 53 Discoverable Tools

Adapter Tools Key Capabilities
Snowflake 6 List databases, schemas, tables, describe tables, run queries
PostgreSQL 6 List DBs, schemas, tables, encryption check, query
AWS 7 S3 buckets, RDS, Glue, Athena, IAM roles, encryption
Azure 7 Storage, SQL DB, Purview assets, Synapse, Key Vault
Jira 6 Projects, issues, compliance tasks, audit trail
Databricks 5 Clusters, jobs, Unity Catalog, SQL, lineage
SQL Server 4 Databases, tables, describe, query
ServiceNow 4 Incidents, CMDB, change requests
Power BI 4 Workspaces, reports, datasets, sources
Purview 4 Search, classifications, lineage, scan status

Architecture

Agent → MCP Registry → Adapter → External System
         │
    ┌────┴────┐
    │ Registry │
    │ 10 types │
    │ CRUD API │
    │ Tool Exec│
    └─────────┘

Knowledge Graph

Graph Model — 29 Nodes, 31 Relationships

Customer Data ──→ Customer 360 ──→ customers ──→ email
                                                  │
                       ┌──────────────────────────┼──────────────────────┐
                       ▼                          ▼                      ▼
                 CDP-ART-34                  GDPR Art.32           CDP-ART-49
                       │                          │                      │
                       ▼                          ▼                      ▼
                 Control-Enc-01                                   Control-CB-03
                       │
                       ▼
                 Unencrypted PII at rest
                 (CRITICAL, score 0.72)

Key Queries

  • Lineage: Trace a column's full governance context (table → product → domain → policy → control → risk)
  • Impact Analysis: Find all assets affected if a policy changes
  • Compliance Gaps: PII columns with no mapped security control
  • Risk Heatmap: Top risks by likelihood×impact score

API Reference

Full Swagger docs at http://localhost:8001/docs

Core Endpoints

Method Path Description
GET /api/health Health check (PG + Neo4j status)
GET /api/workflows List workflows
GET /api/workflows/:id Get workflow with nodes/edges
POST /api/workflows/:id/execute Execute a workflow
GET /api/executions List executions
GET /api/agents List registered agents
GET /api/approvals List pending approvals
POST /api/approvals/:id/approve Approve a checkpoint
WS /ws/workflows/:id Live execution streaming

Dashboard Endpoints

Method Path Description
GET /api/dashboards/mission-control Main dashboard stats
GET /api/dashboards/governance Governance metrics
GET /api/dashboards/security Security findings + risks
GET /api/dashboards/compliance Framework scores + gaps

Knowledge Graph Endpoints

Method Path Description
GET /api/knowledge/summary Graph statistics
GET /api/knowledge/lineage?column=email Full lineage trace
GET /api/knowledge/risks Risk heatmap
GET /api/knowledge/compliance/gaps Unprotected PII columns

MCP Endpoints

Method Path Description
GET /api/mcp/adapters List available adapters
GET /api/mcp/servers List connected servers
POST /api/mcp/servers Connect a new server
POST /api/mcp/tools/execute Execute a tool

Organization Endpoints

Method Path Description
GET /api/organizations/ List organizations
GET /api/organizations/current Current user context
GET /api/organizations/:id/members List members with roles

Reports Endpoints

Method Path Description
GET /api/reports List generated reports
POST /api/reports/generate/executive-summary Generate executive summary
GET /api/reports/:id/download Download as TXT/JSON

Configuration

All configuration via environment variables in backend/.env:

# Server
BACKEND_URL=http://localhost:8001
FRONTEND_URL=http://localhost:3000
HOST=0.0.0.0
PORT=8001

# Databases
DATABASE_URL=postgresql://senanalytics:***@localhost:5432/senanalytics
NEO4J_URI=bolt://localhost:7687
REDIS_URL=redis://localhost:6379

# LLM (optional — without it, agents use simulated data)
LLM_API_KEY=sk-***
LLM_MODEL=gpt-4o-mini

Deployment

Local Development

cd docker && docker compose up -d         # Infrastructure
cd backend && uvicorn app.main:app --port 8001  # Backend
cd frontend && pnpm dev --turbopack -p 3000     # Frontend

Docker Compose

docker compose up -d                      # All services

Kubernetes

kubectl apply -f k8s/                     # Full cluster deploy
kubectl -n senanalytics get pods,svc,ingress

Project Stats

Metric Value
Total files 107
Total lines 14,492
Python 48 files · 5,522 lines
TSX/TS 39 files · 4,118 lines
YAML 13 files · 2,969 lines
Backend routes 47 endpoints
Frontend pages 14 routes
Docker images 2 (backend 389MB, frontend 283MB)
K8s manifests 9 files

Documentation

Document Path
User Guide docs/USER_GUIDE.md
Architecture Diagram docs/architecture-diagram.html
Database Schema docs/database-schema.sql
Knowledge Graph Schema docs/knowledge-graph-schema.md
Event Model & API docs/event-model-api.md

License

Proprietary — Sen'Analytics. Built for data governance in Senegal and Francophone West Africa.

About

Agentic Governance Operating System — Visual AI-powered governance, compliance, security, and risk management platform

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages