Comp AI Architecture Documentation
comp-ai
Comp AI Architecture Documentation
Comp AI is an open-source (AGPLv3) compliance automation platform that helps organizations achieve and maintain SOC 2, ISO 27001, HIPAA, and GDPR compliance. It uses LLMs to automate policy generation, questionnaire answering, vendor risk assessment, and provides an AI-powered chat assistant for compliance guidance.
Documents
| Document | Description |
|---|---|
| 00 - Overview | High-level project summary, tech stack, and monorepo layout |
| 01 - Entry Points | Application entry points, authentication flow, and request lifecycle |
| 02 - Core Logic | Business domain model, data flows, background jobs, and integrations |
| 03 - Key Abstractions | Design patterns, interfaces, and architectural decisions |
| 04 - Configuration | Environment variables, Docker setup, CI/CD, and deployment |
| 05 - AI and LLM | LLM usage, RAG pipeline, embeddings, and AI architecture |
Quick-Reference Tech Stack
| Layer | Technology |
|---|---|
| Language | TypeScript |
| Monorepo | Turborepo + Bun |
| Frontend | Next.js 16, React 19, Tailwind CSS v4 |
| Backend API | NestJS 11 (Express) |
| Database | PostgreSQL + Prisma 6.18 |
| Auth | Better Auth (magic link, OTP, OAuth, JWT) |
| AI/LLM | Vercel AI SDK, OpenAI, Anthropic, Groq |
| Vector Store | Upstash Vector (text-embedding-3-small) |
| Background Jobs | Trigger.dev v4 |
| Cache/KV | Upstash Redis |
| Resend + React Email | |
| File Storage | AWS S3 |
| Analytics | PostHog, Vercel Analytics |
| UI Components | Radix UI, TipTap editor |
| Integrations | AWS, Azure, GCP, GitHub, Google Workspace, JumpCloud, Ramp, Rippling, Vercel, Aikido |
| CI/CD | GitHub Actions, semantic-release |
| Container | Docker multi-stage (Bun build, Node 22 runtime) |