Agentic AI-Powered Multi-Standard Compliance Assessment Platform
Engineered by Vaibhav Verma for Enterprise Core's Hackathon 2026 | Powered by OmniAgent™
SentriX is a production-grade enterprise application that transforms how organizations approach regulatory compliance. Upload your policy documents and receive instant readiness scores, gap analysis, evidence validation, maturity levels, and phased remediation roadmaps across multiple ISO standards simultaneously — powered by a pipeline of specialized AI agents and a 3-tier hybrid scoring engine.
Supported Standards: ISO 37001 (Anti-Bribery) · ISO 37301 (Compliance) · ISO 27001 (InfoSec) · ISO 9001 (Quality) · ISO 37000 (Governance) · ISO 37002 (Whistleblowing)
| Innovation | Description |
|---|---|
| 7-Agent Orchestrated Pipeline | Coordinated team of specialized AI agents that hand off structured context through a sequential pipeline |
| 3-Tier Hybrid Scoring | ML semantic scoring → Groq AI enhancement → keyword+NLP fallback with graceful degradation |
| Evidence Validation | Validates whether cited evidence actually supports compliance claims — checks sufficiency, quality, and cross-standard reuse |
| Policy Generation | Generates compliant policy documents addressing all identified gaps for immediate download |
| Cross-Standard Synergy | Identifies where a single remediation action satisfies requirements across multiple standards |
| Real-Time Streaming | SSE-based live progress as each agent executes |
| OmniAgent Integration | Architected for Enterprise Core's OmniAgent infrastructure with transparent fallback |
- Node.js 18+ and npm 9+
- Anthropic API key (optional — demo mode works without it)
- Python 3.9+ (optional — for ML scoring)
- PostgreSQL 14+ (optional — for persisted assessment sessions/results)
# Server
cd server
npm install
cp .env.example .env # Add your ANTHROPIC_API_KEY and optional DATABASE_URL
npm run dev # → http://localhost:3001
# Client (separate terminal)
cd client
npm install
npm run dev # → http://localhost:5173If you want assessment sessions, logs, uploaded-document metadata, and final results to survive server restarts, set DATABASE_URL in .env:
DATABASE_URL=postgresql://postgres:postgres@localhost:5432/SentriXYou can start a local Postgres instance from the repo root with:
docker compose up -d postgresWhen DATABASE_URL is present, the server bootstraps the required assessment_sessions table automatically on startup. If it is absent, the backend continues using the existing in-memory runtime store.
Click "Try Demo" in the navbar to load a complete sample assessment with all agents' output — no API keys or uploads required.
| Layer | Technologies |
|---|---|
| Frontend | React 19, TypeScript, Vite 6, Tailwind CSS 4, Framer Motion, Recharts, Zustand, React Router 7 |
| Backend | Express 5, TypeScript, Anthropic SDK, Groq SDK, pg, pdf-parse, mammoth, Multer, SSE |
| AI/ML | Anthropic Claude (claude-sonnet-5, claude-fable-5), Groq Cloud (openai/gpt-oss-120b), sentence-transformers (Python), OmniAgent platform |
| Tooling | ESLint, tsx (watch mode), jsPDF, D3, Fuse.js |
Client (React) ──HTTP/SSE──▶ Express API ──▶ AI Pipeline (7 agents)
│ │
├──▶ Claude API ├──▶ Document Parsing
├──▶ Groq Cloud ├──▶ Standard Scoring (×4)
├──▶ OmniAgent ├──▶ Gap Analysis
└──▶ Python ML ├──▶ Evidence Validation
├──▶ Remediation Planning
└──▶ Policy Generation
The pipeline uses a 3-tier provider cascade — OmniAgent → Anthropic Claude → local generation — so every agent produces results regardless of external service availability.
SentriX/
├── client/ # React 19 frontend
│ └── src/
│ ├── components/ # Layout, dashboard, analytics, reports, UI primitives
│ ├── pages/ # 13 route pages
│ ├── store/ # Zustand state management
│ ├── hooks/ # Custom React hooks
│ ├── utils/ # API client, helpers, report generation
│ ├── data/ # Standards data, demo assessment data
│ └── types/ # TypeScript interfaces
├── server/ # Express 5 backend
│ └── src/
│ ├── agents/ # Agent runner + orchestrator
│ ├── services/ # Pipeline, scoring, copilot, GenW bridge
│ ├── routes/ # 8 API route modules
│ ├── data/ # ISO standards, questionnaires, knowledge base
│ └── middleware/ # File upload handling
├── docs/ # Comprehensive documentation (13 sections)
└── .env.example
SentriX includes a comprehensive documentation system covering every aspect of the platform.
| Document | Description |
|---|---|
| 01 — Project Overview | Problem statement, target users, capabilities, supported standards |
| 02 — System Architecture | Layer architecture, data flow diagrams, design decisions |
| 03 — AI Agent Pipeline | All 7 pipeline agents with inputs, processing, outputs, and prompt patterns |
| 04 — Frontend Architecture | Component system, routing, state management, design system |
| 05 — Backend Services | Service catalog, pipeline orchestration, scoring engine, GenW bridge |
| Document | Description |
|---|---|
| 06 — API Documentation | All endpoints with request/response examples |
| 07 — Data Model | Entity definitions, TypeScript interfaces, ER diagrams |
| 08 — Analytics & Scoring | Scoring algorithms, confidence calculation, forecasting |
| Document | Description |
|---|---|
| 09 — Security & Data Handling | Upload controls, data privacy, audit logging, dependency security |
| 10 — Deployment Guide | Local setup, production build, Docker, environment configuration |
| 11 — Future Roadmap | Planned features across 5 phases |
| 12 — Contributing Guide | How to add standards, agents, analytics modules, and endpoints |
| Document | Description |
|---|---|
| Agent Design | Original multi-agent pipeline specification |
| API Reference | Original API endpoint reference |
| Architecture | Original system architecture document |
| Evidence Validation | Evidence Validation Agent specification |
| OmniAgent Integration | OmniAgent platform integration guide |
| Novelty | Value proposition and differentiators |
| Scoring Engine | 3-tier hybrid scoring documentation |
| Wireframes | UI flow wireframes and design tokens |
Proprietary — Built for Enterprise Core Hackathon 2026 by Vaibhav Verma
SentriX™ — Where Integrity Meets Compliance Intelligence