AI Product Engineer · Orange Business

I design AI products and get teams to adopt them.

As a work-study student at Orange Business, on the Private AI Factory team, I scope business needs and build the AI agents and solutions that meet them. Alongside this, I study project management and digital marketing at ISG.

Scroll
Portrait of Dorian Fosse
Logbook

Projects

2 projects · 2 live
Orange Business · 2026 — present · live

Vigie

An agent that reads a team's documents — meeting minutes, specifications, tenders —, answers while citing its sources, then acts in the workspace: project created, task added, meeting scheduled. Live since August 31, 2026, by personal invitation: around ten users to date.

Impact
0 actions taken without human validation, by design
Role
AI Product Engineer
Stack
Agent LLM, React, Node
Read the case study →
Personal · 2026 · live

InfoPC

A system monitor in the menu bar of Apple Silicon Macs: per-core CPU, GPU, memory, disk, network, processes — and above all the actual fan speed, with a slider to force it, something macOS exposes nowhere. Published as open source, installable with a single Homebrew command.

Impact
340 KB for the entire application, without a single external dependency
Role
Design and development
Stack
Swift, SwiftUI, IOKit / SMC
Read the case study →

All projects

The story

I don't come from engineering, and that's precisely the point

I started in management: economics in Montreal, sports management, then a master's in management and project management at ISG. Coding was something I kept from high school — mathematics and computer science — without ever thinking it would become my job.

What changed was the moment language models made building accessible to anyone who knows how to frame a problem. I realised the hard part wasn't writing the code: it was knowing what to build, for whom, and getting it adopted. That, I knew how to do.

Today, at Orange Business, I design agents that run for real users. I scope the need with them, I build, I ship, and I check whether it gets used a second time. Everything else is just a demo.

The method

Three rules, and I break none of them

  1. 01

    Understand the real need

    I start from the actual workflow: who does what, how long it takes, where it gets stuck. The initial request rarely describes the problem, and half the requests can be solved without AI — better to know that early.

  2. 02

    Build fast, build small

    A first usable version in a few days rather than an overhaul in a few months. We fix things against reality, not assumptions. What gets delivered has to run, deploy, and be picked up by someone else.

  3. 03

    Measure what is useful

    The only metric that matters: is the tool used a second time, and by whom. We keep what is useful and cut the rest. A tool that gets abandoned costs more than a tool that was never built.

Skills

Product and engineering, without choosing.

AI

  • LLM agents & tool loops
  • RAG & sourced answers
  • Prompt engineering
  • LLM security (injection, guardrails)
  • Cost & quota control

Development

  • TypeScript · React · Node
  • PostgreSQL
  • Swift · SwiftUI
  • Git & CI/CD
  • Deployment & hosting

Product

  • Scoping business needs
  • Project management
  • Launch & adoption
  • Digital marketing
  • Presentation & demo

Background

Experience
  1. July 2026 — present

    AI Product Engineer — work-study

    Orange Business · Private AI Factory team · Lyon, France

    • Scoping business needs and building the AI agents and solutions deployed in a private environment, from the first conversation through to going live.
    • Design and development of Vigie: a document agent that answers by citing its sources and acts within the workspace, never taking an action without approval.
    • Security and cost control treated as features: invitation-only access, data hosted in Europe, spending cap enforced server-side.
Education
  1. 2026 — 2028

    Master's in Management and Project Management · Digital Marketing track

    ISG · in progress

  2. 2023 — 2026

    Bachelor's in Sports Science (STAPS) · Sports Management

    Université de Perpignan (UPVD)

  3. 2022 — 2023

    B.Sc. in Management Economics

    Université de Montréal

  4. 2018 — 2021

    Scientific Baccalauréat · Mathematics, Computer Science (NSI)

    Lycée International de Saint-Genis-Pouilly

Contact

A project, a question, an opportunity?

I'm on a work-study contract until 2028 and I reply to everyone. Email is the easiest.

Based in
Geneva, Switzerland
Status
Work-study placement at Orange Business until 2028
Languages
Native French, English B2