# yonyon.ai — Yonatan Gross (full)

> AI Platform Engineer & Backend Developer. Builds production AI systems end-to-end: RAG pipelines, multi-agent architectures, MCP servers, infrastructure.

This is the comprehensive companion to [/llms.txt](https://yonyon.ai/llms.txt). It carries the full bio, every project, both service tracks, the chat API contract, an FAQ, and contact details.

## Bio

Yonatan Gross designs and operates production AI systems solo. He built a WhatsApp AI platform that combines RAG retrieval, commerce, and content pipelines into one operated product. He created [OrchestKit](https://github.com/yonatangross/orchestkit), an open-source Claude Code agent framework with 100+ skills, 35+ agents, and 200+ hooks. He advises teams on Claude Code adoption and LLM integration — readiness assessment, architecture, model selection, RAG strategy, and multi-agent design — and ships the infrastructure those systems run on (Terraform, Docker, Cloudflare, GitHub Actions).

## Projects

### [OrchestKit](https://yonyon.ai/projects/orchestkit)

Open-source Claude Code agent framework: 100+ skills, 35+ agents, and 200+ hooks that extend Claude Code into a full development platform. It standardizes how agents explore, plan, implement, and verify work across a repo. Repo: [github.com/yonatangross/orchestkit](https://github.com/yonatangross/orchestkit).

### [AI Operations Platform](https://yonyon.ai/projects/ai-operations-platform)

Production AI built and operated end-to-end: a WhatsApp assistant backed by RAG retrieval, a commerce flow, and content pipelines. Custom MCP servers wire the assistant into the underlying tools and data. Details: [yonyon.ai/projects/ai-operations-platform](https://yonyon.ai/projects/ai-operations-platform).

### [SkillForge](https://yonyon.ai/projects/skillforge)

A multi-agent platform that turns URLs into ready-to-use Claude Code and Cursor artifacts. It crawls a source, extracts the useful structure, and emits skills and configuration agents can consume directly. Details: [yonyon.ai/projects/skillforge](https://yonyon.ai/projects/skillforge).

### [Trading AI Analyst](https://yonyon.ai/projects/trading-ai-analyst)

A multi-agent LangGraph trading platform spanning analysis, signal generation, risk modeling, and real-time streaming. Specialized agents coordinate through a shared graph to produce and pressure-test trading decisions. Details: [yonyon.ai/projects/trading-ai-analyst](https://yonyon.ai/projects/trading-ai-analyst).

## Services

### Build

- **AI Development** — LLM chatbots, RAG / semantic search, multi-agent orchestration, LangGraph workflows, production deployment.
- **Full-Stack Development** — React 19, Next.js 16, FastAPI, PostgreSQL / Redis, Docker & CI/CD.
- **Infrastructure & DevOps** — Terraform, Docker, Cloudflare Tunnels, VPS, GitHub Actions.
- **AI Agent Development** — custom MCP servers, OrchestKit skills / agents, Claude Code tooling.

### Advise

- **AI Consulting** — readiness assessment, architecture, LLM selection, RAG strategy, multi-agent design, implementation roadmap.
- **AI Integration Audit** — review an existing LLM integration for correctness, cost, and reliability.
- **Coaching** — hands-on guidance for engineers adopting Claude Code and LLM workflows.
- **Workshops** — team sessions on Claude Code adoption and production LLM patterns.

## Chat API

A public, no-auth endpoint that answers questions about Yonatan's work. Responses stream as plain text.

- **Endpoint:** `POST` [/api/chat](https://yonyon.ai/api/chat)
- **Auth:** none (public, read-only)
- **Model:** `claude-haiku-4-5`
- **Response:** streaming `text/plain`
- **Rate limit:** 10 requests per IP per minute
- **Body:** `{ "messages": [{ "role": "user", "content": "..." }] }` — at most 20 messages, each `content` at most 2000 characters.

### curl example

```bash
curl -N https://yonyon.ai/api/chat \
  -H 'Content-Type: application/json' \
  -d '{"messages":[{"role":"user","content":"What does Yonatan build?"}]}'
```

See also the [ask-yonatan skill](https://yonyon.ai/.well-known/agent-skills/ask-yonatan/SKILL.md) and the [OpenAPI spec](https://yonyon.ai/.well-known/openapi.json).

## FAQ

**What does Yonatan do?** He designs and operates production AI systems solo — RAG pipelines, multi-agent architectures, MCP servers, and the backend infrastructure they run on — and advises teams on Claude Code adoption and LLM integration.

**How do I reach him?** Email [yonaigross@gmail.com](mailto:yonaigross@gmail.com), or book a free 15-minute intro call at [cal.com/yonatan-hq](https://cal.com/yonatan-hq).

**Can an agent use the API?** Yes. `POST` [/api/chat](https://yonyon.ai/api/chat) is public and unauthenticated, rate-limited to 10 requests per IP per minute, and streams plain text. In-browser agents can also use the WebMCP tools exposed via `navigator.modelContext` (`ask_yonatan`, `book_intro_call`, `browse_projects`).

**Is there pricing?** No pricing is published. Engagements are scoped per-project — [book a call](https://cal.com/yonatan-hq) to discuss scope.

**How is this different from hiring an agency or a freelance marketplace?** You work directly with the senior engineer who designs, builds, and operates the system — no handoffs, no junior staffing, no account-management layer. Decisions and accountability sit with one person from architecture to production, and iteration is faster because nothing is lost in translation. The honest trade-off: one engineer is not a fit when you need a ten-person delivery team staffed overnight; for that, an agency is the right call.

## Contact & sameAs

- Email: [yonaigross@gmail.com](mailto:yonaigross@gmail.com)
- GitHub: [github.com/yonatangross](https://github.com/yonatangross)
- LinkedIn: [linkedin.com/in/yonatangross](https://linkedin.com/in/yonatangross)
- X: [x.com/yonyoniz](https://x.com/yonyoniz)
- Book a call: [cal.com/yonatan-hq](https://cal.com/yonatan-hq)