ML/AI Lead Engineer · Forward Deployed AI Engineer · Python · AWS · Agentic Systems
📍 Based in Switzerland · 🌍 Open to Full Remote Opportunities
📧 patricia@wintrebert.ch
🔗 LinkedIn · GitHub
AI engineer and technical lead specializing in RAG systems, agentic workflows, and production AI deployment. I design and ship end-to-end AI solutions — from data ingestion and vector indexing to LLM orchestration and cloud deployment — in fast-paced, startup-style environments.
I bridge the gap between business needs and technical execution: hands-on in the code, fluent with stakeholders, and experienced leading remote development teams across the full delivery lifecycle.
Trilingual (English, French, German) and comfortable in multicultural, international contexts.
- RAG pipelines, vector databases (Qdrant, pgvector), embeddings (OpenAI, Azure OpenAI)
- LLM orchestration, multi-agent systems, MCP-compatible agentic workflows
- Claude API, OpenAI API, LangChain, NLP, Generative AI
- Python, FastAPI, Pydantic
- AWS (Lambda, App Runner, ECS, ECR, Amplify, RDS)
- Azure (OpenAI, Speech)
- Docker, GitHub Actions CI/CD, Supabase
- MLOps, production monitoring
- Next.js 14, React, TypeScript, Tailwind CSS
- REST APIs, Supabase auth, RLS
- Agile / Scrum, sprint coordination, remote team management
- Stakeholder workshops, business-to-architecture translation
- End-to-end deployment ownership
ML/AI Lead Engineer & Forward Deployed AI Engineer · Remote · Contract
data IQ AG is a Swiss AI consulting agency specialized in Market Research, Qualitative Research, and Customer Experience. Insight Lab AI is its flagship SaaS platform for the full qualitative research lifecycle — transcription, anonymisation, and AI-assisted reporting.
What I shipped in production:
- Designed and owned end-to-end delivery of RAG pipelines and AI agents: data ingestion, embeddings, vector storage (Qdrant + pgvector), API serving via AWS Lambda
- Built agentic workflows integrating LLMs (OpenAI, Claude) with client data and internal APIs — MCP-compatible orchestration layer for multi-step agent execution
- Built a custom RAG system for a major German market research firm: ingested qual discussion guides + quant questionnaires, indexed vectors with metadata in Qdrant, FastAPI backend on AWS App Runner (ECR + RDS PostgreSQL), cosine similarity search with multi-filter support
- Built and deployed a custom qualitative survey platform for consulting operations: conversational flow, AI-driven probe generation (Azure OpenAI + probabilistic topic extraction), Azure Speech transcription, FastAPI on AWS ECS, Next.js on Amplify, Supabase with RLS
- Built and operate a production multi-agent coding orchestration system: Claude Code across isolated git worktrees, Notion as operational source of truth, GitHub Actions CI, automated cost quota guard, persistent agent knowledge base for durable cross-session context
- Contributed to Insight Lab AI core development: project management, sprint coordination, hands-on development, CI/CD ownership
- Supervised developers, led agile sprints, owned deployments end-to-end
- Conducted technical workshops; translated client needs into concrete architecture
Stack: Python · FastAPI · Next.js · React · TypeScript · AWS Lambda · App Runner · ECS · Amplify · Qdrant · pgvector · OpenAI API · Claude API · Supabase · RAG · Docker · GitHub Actions · Agile
AI Engineer
- Contributed to development of AI-powered tools for conversational data analysis
ML Engineer
- Participated in applied machine learning projects supporting women in tech initiatives
Production agentic infrastructure for automated software development: deterministic Python scripts orchestrate Claude Code across isolated git worktrees, with Notion as operational brain, GitHub Actions CI, automated cost quota guard, and a persistent knowledge base (incident post-mortems, code maps, architecture decisions) for durable cross-session agent context.
Dockerized dashboard for discovering and ranking US small-cap stocks. Discovers candidates from NASDAQ and Finviz, enriches with yfinance, applies hard filters, computes a setup score, and exposes results via FastAPI + React/Vite frontend.
🔗 github.com/patw47/smallcaps-screener
Automated stock monitoring bot built on n8n + Claude AI. Sends a daily briefing via Telegram and Gmail for each tracked ticker.
🔗 github.com/patw47/stock-tracker
Institutional-style cognitive trading architecture combining real-time market intelligence, historical pattern analysis, structured LLM reasoning (OpenClaw agents), and a trade journal system for behavioral self-review and adaptive learning.
💬 Additional private projects available upon request.
- 🎓 Mines ParisTech – PSL, France — Master's in Machine Learning Engineering (2023)
- 🎓 University of Strasbourg, France — Bachelor's in Computer Science (2013)
- 🥇 Winner — Deploy Impact Hackathon, Zurich (2023)
- 🥈 2nd Place — Hack'n'Lead Hackathon, Zurich (2023)
- 🧠 Member of Mensa Switzerland & Intertel
Algorithmic and discretionary trading on Forex and indices · AI-assisted trading strategies · Market psychology, liquidity zones, volatility dynamics · Crypto and blockchain
🇬🇧 English — Fluent · 🇫🇷 French — Native · 🇩🇪 German — Fluent