Every employee gets a personal AI agent. The agents talk to each other. Work that currently requires human chains happens instantly.
# 1. Copy environment variables
cp .env.example .env
# Fill in your ANTHROPIC_API_KEY and ETO_API_KEY
# 2. Start infrastructure (Postgres + Redis)
docker compose up -d
# 3. Start the API server
cd packages/api
pip install -e ".[dev]"
uvicorn main:app --reload --port 8000
# 4. Start the web app (separate terminal)
cd packages/web
npm install
npm run devAPI docs: http://localhost:8000/docs
e-agent/
├── packages/
│ ├── api/ # FastAPI backend (Python)
│ │ ├── agents/ # Agent runtime + Claude integration
│ │ ├── messaging/ # ETO inter-agent communication
│ │ ├── registry/ # Agent discovery service
│ │ ├── approvals/ # Human-in-the-loop approval flows
│ │ ├── permissions/ # RBAC + IdP sync
│ │ ├── admin/ # Admin Control Plane (Phase 2)
│ │ └── integrations/ # SaaS connectors (Phase 1+)
│ ├── web/ # React/Next.js frontend (TypeScript)
│ │ └── src/
│ │ ├── app/ # Pages
│ │ ├── components/ # Chat UI, approval pop-ups
│ │ └── lib/ # API client
│ └── shared/ # Shared types and protocol definitions
├── infra/ # Docker, Fly.io deploy configs
├── scripts/ # Dev tooling, seed data
└── docs/ # Architecture decisions
| Layer | Choice |
|---|---|
| Agent runtime | Python (FastAPI) + Claude API |
| LLM | Claude (Haiku for routing, Sonnet for reasoning) |
| Frontend | TypeScript, React, Next.js, Tailwind |
| Inter-agent comms | ETO (messaging + payments + file transfer) |
| Database | PostgreSQL + pgvector |
| Cache/queue | Redis |
| Infra | Docker, Fly.io |
- Muhammad — CTO
- Max — CEO
- Likitha — COO
See IMPLEMENTATION_PLAN.md for the week-by-week build plan.