Edouard Baillot
AI product engineer — LLM agents in production
Open to contract and full-time roles
Product engineer specialized in LLM agent systems in production: I designed and operate an autonomous multi-agent infrastructure that builds and maintains my four live SaaS products. Enterprise engineering by day at La Poste / Colissimo. Before that, ten years in the French Foreign Legion.
Experience
2025 – Present
Founder & engineer — agent-operated SaaS portfolio
DesignPilot · Rat List · Teste Nacionalidade · Recevance
Claude API · MCP · Next.js · TypeScript · Supabase · PostgreSQL · Stripe · Redis · Docker / VPS · Vercel
- Designed and operate an autonomous multi-agent infrastructure that maintains four live products: LLM agents plan, code, review, and ship through a git-based state layer on three machines, with human approval gates on everything outward-facing.
- Engineered reliability by measurement: found a 35% violation rate on prompt-only rules and moved critical invariants into deterministic tool-layer guards (pre-execution hooks, deny policies) — agents are audited, not trusted.
- Route work across model tiers by task class to balance cost and quality; token budgets and scheduled routines keep the system running unattended.
- Ship AI features end to end: agents that turn emails and voice notes into structured tasks, assisted search, a daily content pipeline (draft → human approval → publish), and a companion mobile app on the App Store and Google Play.
2025 – Present
Full-stack engineer
La Poste / Colissimo
Java · Spring Boot · Angular · Oracle
- Deliver e-invoicing compliance across internal flows: business rules, document generation, and shipment tracking.
- Extend and maintain production applications used across the company — dashboards, monitoring, and parcel management.
- Migrate legacy modules incrementally, covered by integration tests and multi-environment CI/CD.
2020 – 2025
Software engineer — IT department (CORSIC)
French Foreign Legion
PHP → React · Docker · Linux · Grafana
- Rebuilt an internal logistics application from PHP to React on a REST API, cutting user errors by 40%.
- Cut average ticket resolution from two days to four hours across 300+ workstations.
- Trained 400+ users on security practices; Docker CI/CD pipeline (−60% release time), Grafana / Prometheus monitoring.
2015 – 2020
Operational unit
French Foreign Legion
- Served five years in a combat unit — discipline, decisions under pressure, and teamwork where mistakes are not an option.
Skills
- AI
- Claude API / Anthropic stack · Multi-agent orchestration · MCP · Guardrails & human-in-the-loop · Prompt engineering · Content pipelines
- Frontend
- TypeScript · React / Next.js · Angular · SCSS
- Backend
- Node.js · Java / Spring Boot · PostgreSQL / Oracle · Redis
- Infrastructure
- Docker · CI/CD (GitHub Actions) · Linux / VPS · Vercel
Education
- 2023 – 2027École 42, Paris — software engineering; RNCP Level 7 (master's equivalent), in progress
- 2020CESI, Paris — computer science
- 2014Professional pastry certification (CAP)
Languages
- Frenchfluent
- Russiannative
- Englishprofessional working proficiency