Senior Software Engineer
I design, code and ship AI-driven products. From scoping to production, in a single pair of hands.
- Agents
- LLMs
- MVP
- RAG
- TypeScript
- Next.js
- React
- Postgres
- Stripe
- Node
- Nest.js
- Tailwind
- Product design
- CMS
- GSap
Six ways
to work together.
AI-first web products
Concept to production. From the homepage to agents. Design + code + infra.
MVP shipped in 4 to 6 weeks
Idea to validated product. Tight scope, clean code, in production from week one.
Mobile apps iOS + Android, one codebase
React Native and Expo. Shipped to both stores from a TypeScript codebase shared with the web.
Polished interfaces, fast front-end
Design systems, animations, 60 fps. Pixel-perfect.
LLM agents & workflows
RAG, tool-use, multi-agent, guardrails, continuous evals, costs under control.
Product scoping + mentoring
I ask the real question, then help you answer it. Workshops, handover, mentoring.
What my clients say.
Yet another successful mission with Pierre.
Pierre was highly efficient. We especially appreciated the progress updates he shared and the communication he maintained throughout.
Working with Pierre is a real pleasure. He combines warmth with skill and does everything to deliver on the mission entrusted to him. Highly recommended.
Serious, committed, methodical, with a strong technical background. He's someone who will deliver your projects with real professionalism. Highly recommended.
The questions I ask before writing code.
The ask is rarely
the right question.
"We want an AI chatbot" often means "we want to reduce support time". The first path ships a chatbot; the second, sometimes a better-structured form. I never start without reformulating.
The tools
I run every day.
$ tree ~/ai
├─ models/ │ ├─ claude-opus-5 ● HOT │ ├─ claude-sonnet-5 │ ├─ gpt-5.6 sol / terra │ ├─ muse-glimmer (local) │ └─ fine-tuning (client) ├─ frameworks/ │ ├─ mastra · tools ● HOT │ ├─ vercel ai sdk │ └─ langgraph.js ├─ memory/ │ ├─ postgres + pgvector │ └─ redis (cache) └─ tool-calling/ ├─ mcp ● HOT ├─ function calling └─ structured output
$ tree ~/backend
├─ runtime/ │ ├─ node 22 ● HOT │ └─ bun (edge) ├─ api/ │ ├─ nestjs ● HOT │ ├─ hono · edge │ └─ mastra · tools ├─ db/ │ ├─ postgres + drizzle │ ├─ typeorm │ ├─ prisma │ └─ redis └─ infra/ ├─ docker ├─ vercel · modal └─ github actions
$ cat frontend.json
{ "framework": "next.js 16", "lang": "typescript", "ui": ["tailwind", "motion", "radix"], "animation": ["motion", "gsap"], "data": ["redux", "tanstack"], "design": ["figma", "lottie"] } ✓ stack chosen per project. ✓ no religion, just fit.
$ tree ~/mobile
├─ runtime/ │ ├─ expo sdk ● HOT │ ├─ react native │ └─ expo router ├─ native/ │ ├─ notifications │ ├─ camera · files │ └─ secure storage ├─ build/ │ ├─ eas build ● HOT │ ├─ eas update (ota) │ └─ app store · play store └─ shared/ ├─ typescript ● HOT ├─ api · types └─ design tokens
How I work,
in five beats.
Product scoping
A 45-minute call. We find the real question behind the ask.
Proof of value
5 to 10 days for a priced prototype. We stop here if nothing proves.
Build in sprints
Weekly deliveries, live demos, course-correction as we go.
Ship to production
Infra, evals, monitoring, runbook. You take the wheel with confidence.
Handover
Clean docs, code review, handover. Your team keeps going.
Project
in mind?
A first call, no commitment. We look at your project together, then you decide.
$ book a call