What does Yonatan Gross do?

Yonatan is an AI Platform Engineer & Backend Developer who designs and operates production AI systems solo — RAG pipelines, multi-agent architectures, MCP servers, and the infrastructure that runs them. He built a WhatsApp AI platform and created OrchestKit, an open-source Claude Code agent framework (111 skills, 37 agents, 211 hooks).

How can I reach Yonatan?

Book a free 15-minute intro call at https://cal.com/yonatan-hq, or email yonaigross@gmail.com.

Can an AI agent use the chat API on this site?

Yes. POST to https://yonyon.ai/api/chat with JSON body {"messages":[{"role":"user","content":"..."}]} (up to 20 messages, each up to 2000 chars). Public, no auth, rate-limited to 10 requests per IP per minute, backed by the claude-haiku-4-5 model, streaming a text/plain response. See also https://yonyon.ai/llms.txt and https://yonyon.ai/.well-known/agent-skills/index.json.

How much does it cost to work with Yonatan, and how do I get a quote?

Engagements are scoped per project, so there is no fixed price list — quotes depend on what you are building. The fastest path to a quote is to book a free 15-minute intro call at https://cal.com/yonatan-hq, describe the work, and get a scoped proposal. You can also email yonaigross@gmail.com.

Why choose Yonatan over an agency or a freelance marketplace?

You work directly with the engineer who designs, builds, and operates the system — no account managers, no hand-offs, no junior bench. Yonatan ships production AI end-to-end (RAG, multi-agent, MCP servers, and the infrastructure they run on) as a single accountable point of contact, and has done so in production (a WhatsApp AI platform) and in the open (OrchestKit). For teams that want senior, hands-on ownership rather than a staffing layer, that directness is the difference.

I build production AI systems

RAG pipelines, multi-agent systems, custom MCP servers, and the Terraform-managed infrastructure they run on. Designed, shipped, and operated end to end.

Last shipped: OrchestKit v8.55.2
Working in: Next.js 16 · LangGraph · pgvector
Next opening: June 2026

Book a free 15-min intro call See My Work

TypeScriptPythonNext.jsReactFastAPILangGraphClaude APIOpenAIPostgreSQLpgvectorRedisLangfuseSanity CMSTailwind CSSRadix UIStorybookSQLAlchemyPydanticDockerTerraformAWSPlaywrightVitestMCP ServersRAGCeleryPrometheusVoyageAITurborepoZustandTypeScriptPythonNext.jsReactFastAPILangGraphClaude APIOpenAIPostgreSQLpgvectorRedisLangfuseSanity CMSTailwind CSSRadix UIStorybookSQLAlchemyPydanticDockerTerraformAWSPlaywrightVitestMCP ServersRAGCeleryPrometheusVoyageAITurborepoZustand

Shipped with

Services

What I Do

AI consulting, development, and full-stack engineering

Build

AI Development

Build production-ready AI features for your application. From conversational chatbots to multi-agent systems and RAG implementations.

LLM-Powered ChatbotsRAG / Semantic SearchMulti-Agent OrchestrationLangGraph WorkflowsAPI IntegrationProduction Deployment

Learn more

Full-Stack Development

Modern web applications built with React, Next.js, and Python backends. From MVP to production, with clean architecture and CI/CD.

React 19 / Next.js 16FastAPI / Node.jsPostgreSQL / RedisDocker & CI/CDVercel DeploymentTypeScript End-to-End

Learn more

Infrastructure & DevOps

Production infrastructure with Terraform IaC, Docker deployments, and automated CI/CD. From VPS setup to Cloudflare Tunnels and GitHub Actions.

Terraform IaCDocker DeploymentsCloudflare TunnelsVPS SetupCI/CD PipelinesGitHub Actions

Learn more

AI Agent Development

Custom AI agents and tooling for Claude Code and other platforms. MCP servers, OrchestKit skills, prompt engineering, and workflow automation.

Custom MCP ServersOrchestKit Skills/AgentsClaude Code ToolingAgent FrameworksPrompt EngineeringWorkflow Automation

Learn more

Advise & Coach

AI Consulting

Get expert guidance on integrating AI into your product. I help you identify opportunities, design the right architecture, and create an implementation roadmap.

AI Readiness AssessmentArchitecture DesignLLM Selection & EvaluationRAG Pipeline StrategyMulti-Agent DesignImplementation Roadmap

Learn more

AI Integration Audit

Comprehensive assessment of your current development workflow and business processes to identify high-impact AI integration opportunities. Delivered as a prioritized roadmap.

Discovery session with your teamWorkflow and toolchain analysisAI opportunity identification and scoringPrioritized implementation roadmapEffort sizing per opportunityPresentation of findings

Learn more

AI Development Coaching

Hands-on coaching to level up your AI development — Claude Code, Cursor, MCP servers, prompt engineering, and shipping production AI features. Tailored to where you are; we shape the format together on a quick intro call.

Claude Code & Cursor workflowsMCP servers & toolingPrompt engineeringRAG & multi-agent patternsShipping AI features to productionSupport between sessions

Learn more

Team Workshops

Bring your team up to speed on AI-powered development — live demos, hands-on labs on your codebase, and real workflow integration. Format and depth scoped to your team on an intro call.

Live demos & hands-on labsTool setup for every participantExercises on your codebaseTeam workflow integrationCustom AI agent designPost-workshop implementation guide

Learn more

Why solo

Why hire one engineer instead of an agency?

Solo senior engineer vs agency vs marketplace, on the axes that decide outcomes.

Code authorship

The person you talk to writes the code

Accountability

One name, end to end

Iteration speed

Decision to deploy in hours, not sprints

See the full comparison

Work

Selected Work

Production systems I designed, built, and shipped — from AI platforms to open-source tools

Production · WhatsApp AI live · operated solo since 2025

AI Operations Platform

B2B

Production AI platform — WhatsApp AI assistant with RAG, commerce infrastructure, content pipelines, LLM evaluation, custom MCP servers.

Results

WhatsApp AI assistant with RAG
Commerce + content pipelines
Custom MCP servers

PythonFastAPINext.jsPostgreSQLpgvectorCeleryRedisDocker

View Case Study

Problem

Small businesses need AI-powered customer interaction, commerce, and content automation but can't afford enterprise platforms or dedicated teams to build and maintain them.

Solution

A single integrated platform combining WhatsApp AI (RAG-powered), commerce infrastructure, content pipelines, and LLM evaluation — all deployed on Terraform-managed infrastructure and operated solo.

Live · <50ms anomaly alerts · TASE/NYSE arbitrage

Trading AI Analyst

B2C

Multi-agent trading platform with LangGraph — market analysis, signal generation, risk modeling, portfolio management, and real-time WebSocket streaming.

Results

Multi-agent LangGraph architecture
Real-time signal streaming via WebSockets

PythonFastAPILangGraphPostgreSQLPGVector+4 more

Full-stack AI-powered trading analyst built with a LangGraph multi-agent architecture. Backend services include market data aggregation, AI-driven signal generation, anomaly detection, correlation discovery, historical pattern matching, risk modeling, portfolio management, and a news/social intelligence pipeline. Real-time data streams via WebSockets. Frontend features unified dashboard, strategy management, trading journal, and learning modules.

Open Source · GitHub-starred · 111 skills, 37 agents

OrchestKit

Open Source193

Open-source agent framework for Claude Code — 111 skills, 37 agents, 211 hooks.

Results

111 skills, 37 agents, 211 hooks
Open-source agent framework

TypeScriptClaude CodeAI AgentsesbuildBash

Code

Production-ready plugin system for Claude Code featuring specialized AI agents, reusable skill modules, and lifecycle automation hooks. Covers FastAPI, React 19, LangGraph, security, testing, and more across domain-specific plugins.

Paused · 3-tier R&D · 20+ specialized agents

SkillForge

PrivatePaused

3-tier multi-agent platform (20+ agents) that transforms URLs into implementation-ready artifacts for AI coding assistants.

Results

20+ agents across 3 tiers
3 analysis modes (Quick/Standard/Deep)

PythonFastAPILangGraphPostgreSQLPGVector+3 more

Submit any URL and a tiered agent system analyzes it: Tier 1 (4 agents) extracts insights, Tier 2 (12 agents) validates with tools like Tavily search and npm/PyPI APIs, Tier 3 (4 agents) does deep research with memory. Generates Claude Code/Cursor-optimized artifacts. Includes Socratic tutoring (8 tutor agents) and semantic search library.

Healthcare · Hebrew voice→chart · WhatsApp notify

Voice-to-Chart Clinical AI

B2B

AI sidecar for veterinary clinics — transcribes Hebrew visit audio, extracts structured medical findings, generates copy-paste-ready chart text, and WhatsApps a summary to the pet owner.

Results

Hebrew voice → structured chart
Designed for ~2,400 visits/mo

PythonFastAPILangGraphGoogle GeminiGroq+5 more

View Case Study

Problem

Vets document while examining — eyes on the screen, not the animal — producing rushed, incomplete records under a heavy daily caseload.

Solution

Record the visit and let a LangGraph pipeline transcribe, extract structured findings, format text for the clinic system, and notify the owner over WhatsApp — designed for a clinic handling around 2,400 visits a month at about one cent per visit.

Client CRM · WhatsApp + email · AI drafts, human-approved

Real-Estate Connector CRM

B2B

A personal CRM for a cross-border real-estate connector — a unified WhatsApp and email inbox where AI drafts replies for human approval, with voice transcription, three-language translation, AI summaries, semantic search, and a deal pipeline.

Results

Unified WhatsApp + email inbox
AI-drafted, human-approved replies

Next.jsReact 19Supabase (PostgreSQL + pgvector)Anthropic (Claude)OpenAI+4 more

View Case Study

Problem

A solo connector juggling buyers across countries loses context across WhatsApp, email, voice notes, and documents in three languages — and can't risk an AI replying to clients unchecked.

Solution

Unify every channel into one inbox, draft each reply with AI for human approval, transcribe and translate across Portuguese, English, and Hebrew, and track every deal through a pipeline — with semantic search over the whole history.

Production-ready · PT→EN/HE RAG · LangGraph

Multilingual Document Intelligence

Internal

An AI document platform — on-demand Portuguese→English/Hebrew translation, multilingual semantic search, and RAG-powered Q&A over property documents, with an adaptive UI that picks one of five panel layouts based on context.

Results

PT→EN/HE on-demand translation
Multilingual semantic search + RAG Q&A

PythonFastAPINext.jsPostgreSQLpgvector+5 more

View Case Study

Problem

Cross-border property paperwork arrives in a language you don't read, scattered across PDFs, images, and DOCX — with no way to search or ask questions across it.

Solution

Translate on demand, search semantically across every language at once, and ground a RAG assistant in the full corpus — with an adaptive UI that surfaces the right panel for each task.

Research · LLM reliability scoring · tiered eval

Journalistic Accuracy Scoring

Internal

A pipeline that tracks and scores journalistic reliability, running a LangGraph LLM stack with tiered evaluation — mocked LLMs for fast CI, real cloud models on a nightly cadence.

Results

Cloud-LLM reliability scoring
Two-tier eval (mocked CI + real nightly)

PythonFastAPIAnthropic (Claude)OpenAILangGraph+2 more

View Case Study

Problem

Reporter reliability is asserted, not measured — there is no systematic, repeatable way to track and score accuracy over time.

Solution

Score reliability with a two-tier evaluation pipeline — mocked LLMs for CI speed and real cloud models nightly for behavioral fidelity — and track each source's score over time.

Community · junior-tech jobs · warm-path graph

Network-Powered Job Discovery

B2C

A network-powered job-discovery engine for junior tech professionals — surfaces junior roles and shows community members' warm paths into the hiring organizations.

Results

Junior-role aggregation
Warm-path intro graph

Next.js 16React 19FastAPICelerySQLAlchemy 2.0+6 more

View Case Study

Problem

Juniors are filtered out of cold applications; the roles that actually convert come through warm intros they can't see.

Solution

Aggregate junior roles and map each to community members who can provide a warm introduction — turning a cold job board into a navigable relationship graph.

What Clients Say

Feedback from people I've worked with

“Yonatan built my entire online presence from scratch — portfolio site, SEO, and booking system. He understood my needs as a therapist and created something that truly represents my practice. Professional, fast, and always available.”

Ayala AttiaClinical Emotional Therapist, Private Practice

01 / 04

Motion & Video Art — Making-of

From proof to published

One reel, iterated in the open: a Remotion pipeline proof → voiceover → character-lock → graded cuts, landing on the shipped version. Scroll the timeline to watch the craft compound — tap any version to play it.

Cleanup Crew

7 versions

Veo 3.1
Seedance
Remotion
music-bed
voiceover
character-lock

Step 1
14s
Remotion proof
First end-to-end Remotion composite — proving the pipeline.
Step 2
14s
+ Voiceover
Scripted voiceover track added over the cut.
Step 3
13s
Character-lock
Crew character held consistent across every shot.
Step 4
16s
v5
Extended to 16s with an audio-mix pass.
Step 5
18s
Finale cut
18s multi-shot finale stitched and graded in Remotion.
Step 6
23s
v10
Fuller 23s cut — more beats, tighter music sync.
Step 7
22s★ Published
v14
The shipped cut — final grade, timing and mix.

Horizontal timeline — scroll to watch each version evolve. The published cut is highlighted.

Building in Public

Latest Posts

Deep-dives into AI agent architectures, RAG optimization, and lessons from shipping production AI systems

New posts as they ship

Try Playground

Follow on LinkedIn

About

About Me

Yonatan Gross - AI Platform Engineer & Backend Developer

Full-stack and AI developer based in Israel. I spent years building high-scale financial systems, and since 2025 I design and operate AI platforms solo — from idea to production.

I built a complete AI platform with a smart WhatsApp assistant, commerce system, content management, and custom MCP servers — all running on managed cloud infrastructure.

I created OrchestKit, an open-source project that helps developers work more effectively with Claude Code. I also help businesses integrate AI and smart tooling into their workflows.

Core Technologies

PythonFastAPIClaude Agent SDKMCP ServersRAG / pgvectorTerraformDockerTypeScriptReact 19Next.js 16CeleryLangGraph

Contact

Let's Work Together

Book a free 15-minute call or send me a message.

Book a Free Call

15-minute intro call — no commitment

15 min

June 2026

Sun

Mon

Tue

Wed

Thu

Fri

Sat

Or reach out directly through any of these channels.

[email protected]

X / Twitter

@yonyoniz

GitHub

@yonatangross

Or chat with my AI assistant using the button in the bottom right corner.

I build production AI systems

Shipped with

What I Do

Build

AI Development

Full-Stack Development

Infrastructure & DevOps

AI Agent Development

Advise & Coach

AI Consulting

AI Integration Audit

AI Development Coaching

Team Workshops

Why hire one engineer instead of an agency?

Selected Work

AI Operations Platform

Results

Problem

Solution

Trading AI Analyst

Results

OrchestKit

Results

SkillForge

Results

Voice-to-Chart Clinical AI

Results

Problem

Solution

Real-Estate Connector CRM

Results

Problem

Solution

Multilingual Document Intelligence

Results

Problem

Solution

Journalistic Accuracy Scoring

Results

Problem

Solution

Network-Powered Job Discovery

Results

Problem

Solution

What Clients Say

From proof to published

Remotion proof

+ Voiceover

Character-lock

v5

Finale cut

v10

v14

Latest Posts

About Me

AI Engineering Insights

Let's Work Together

Book a Free Call

What I Do

Build

AI Development

Full-Stack Development

Infrastructure & DevOps

AI Agent Development

Advise & Coach

AI Consulting

AI Integration Audit

AI Development Coaching

Team Workshops

Why hire one engineer instead of an agency?

Selected Work

AI Operations Platform

Results

Problem

Solution

Trading AI Analyst

Results

OrchestKit

Results