Software Engineer
Forward Deployed Engineer (Junior to Confirmed)
Paris
Apply on the employer's siteRole description
Role title:
Forward Deployed Engineer (FDE)
Also known as
Forward Deployed Software Engineer
- Deployment Engineer
- the person who makes software actually work in a factory
Company:
The FDE Company
Location:
Paris-based, deployed at industrial client sites across France and Europe
Contract:
CDI
Level:
Junior (0–3 years of experience). We hire for trajectory, not track record.
Salary range:
€52–62K base + up to €8K deployment bonus tied to validated client gains. Calibrated on 2026 Paris FDE market data (junior band €52–68K).
Training:
3-week paid intensive bootcamp before first deployment — business, client craft, and systems
Start date:
Rolling cohorts from Q4 2026
About Us
Our mission.
We make industrial AI software deliver the value it promises — measured in euros on the client's P&L, not in slides.
Our story.
The FDE Company was founded in 2026 by Devoteam, one of Europe's leading technology consultancies, and OSS Ventures, the studio behind Europe's largest portfolio of industrial AI software. That's the last time you'll hear about our shareholders — from here on, this is about us. Our founding team learned something expensive across more than 100 industrial AI deployments: roughly 40% of the data an AI system needs does not exist in any system of record. It lives in a planner's head, in a supervisor's spreadsheet, in the muscle memory of a machine operator with twenty years on the line. Someone has to go get it. That someone is you.
Our ambition.
The industry data is brutal: about 95% of enterprise AI pilots produce no measurable business impact (MIT, 2025). Not because the models are weak — because nobody owns the deployment. Palantir built a multi-billion-dollar company by owning it with Forward Deployed Engineers. OpenAI and Anthropic both launched deployment ventures in 2026 for the same reason. We are building the European industrial version: a trained corps of FDEs who embed inside factories and supply chains, deploy proven vertical AI software, and stay until the gains are validated by the client's CFO.
Our track record, inherited from day one.
You will not be deploying prototypes. The method you'll be trained on produced several million dollars in CFO-validated recurring gains at a US plastics manufacturer, and cut software deployment time by 10x at a global sporting-goods company. Deployment cycles here are measured in weeks where the industry measures in years.
Our values.
These are filters for fit, not wall decorations.
- Impact, not reports. Every hour of work traces to the client's P&L. If you can't draw the line from your task to their competitiveness, question the task.
- Ship in weeks. Velocity is direction times speed. A working system on Friday beats a perfect architecture next quarter, because action breeds information.
- The floor is the truth. The ERP says one thing, the shop floor does another. We believe the floor, and we go see it.
- Ownership without borders. Problems travel fast here. You close gaps even when they're outside your lane. No submarines.
- Humility as a method. "I don't understand this process yet" is the beginning of every good deployment. Scout mentality: the goal is to see what's true, not to be right.
Why this window won't stay open
Industrial AI deployment is already happening — at scale, with validated euros — in the US. Palantir built its playbook there. OpenAI and Anthropic both launched deployment arms in 2026 because they saw the same thing we did. The plant we mention above, the one with several million dollars in CFO-validated gains, is American. Europe will get there too — it always does, about five years later, once the method has been de-risked, translated, and stripped of whatever made it move fast in the first place. That's the normal lag. This role skips it. We're not importing a watered-down, "Europeanized" version of what's already running in US plants — we're bringing the method here directly, now, while the French market still thinks this is five years away. Join early enough and you're not learning a trade that already exists here; you're one of the people who makes it exist.
Why we're hiring for this role
Demand exists; the bench does not. Engineers who can hold a room with a plant director in the morning and ship production code in the afternoon are the rarest profile in French B2B software right now, and the market knows it — FDE job postings grew roughly 800% in 2025 alone, and almost nobody in Europe trains this profile deliberately.
We stopped waiting for these people to appear on the market. We are building them. That is the founding act of The FDE Company: hire high-slope juniors, train them hard for three weeks, embed them alongside senior deployment leads, and grow the corps that industrial AI in Europe is missing. You are the first cohorts. The playbook you help write becomes the standard for everyone after you.
What You'll Do
In a nutshell.
You embed inside an industrial client — a factory, a warehouse network, a supply chain organization — and you deploy vertical AI software until it produces measurable, recurring gains. You own the last mile that everyone else abandons: from signed contract to validated euros.
Your Main Responsibilities
- Deploy vertical AI products for industrial operations — production planning and scheduling, product data intelligence, workforce orchestration, shop-floor collaboration, supply chain planning, procurement automation — inside client environments: configuration, integration, data pipelines, and the custom glue code no product ships with.
- Extract the data that doesn't exist yet. Interview the planner, sit with the operator, dig through the shared drive, and turn tribal knowledge into structured, versioned data the system can run on.
- Frame every scope as a business case: what process changes, what it's worth in euros, how we'll measure it, and when the client's finance team signs off on the number.
- Run the deployment cycle like a professional: weekly steering with client stakeholders, clear status, risks surfaced early, decisions documented — the discipline of a top-tier consulting project with the output of an engineering team.
- Feed the field back into the product. What you learn on the floor becomes roadmap input for the software partner whose product you deploy. You are their eyes in production.
On a day-to-day basis, you will
- Walk a production line with a supervisor and map how the real process differs from the documented one.
- Write a Python connector to a 15-year-old ERP export, clean the data, and get it flowing into the product by end of day.
- Rebuild a plant's scheduling economics in a model the ops director actually trusts, then defend it in front of their finance controller.
- Turn a vague complaint ("the planning is always wrong") into a scoped problem with a success criterion, a data requirement, and a two-week delivery plan.
- Present progress to a room that includes a plant director, an IT manager who didn't ask for you, and an operator who knows more than everyone — and get all three to yes.
Who You'll Work With
You report to a Senior FDE / Deployment Lead who owns the account and reviews your work daily in your first months. You work alongside the product and engineering teams of the software partners whose products you deploy — they are your product line and your escalation path. On the client side, you work with plant directors, supply chain managers, planners, quality engineers, IT teams, and the operators who will use what you build. Behind you: delivery infrastructure at European scale, and a deployment method refined across 100+ industrial AI deployments.
What Success Looks Like
At 30 days (end of bootcamp + first embed)
— you've completed the three-week training, you can read an industrial P&L and explain where your deployment touches it, and you're embedded on your first client site shadowing a senior FDE with real tasks in the sprint.
At 90 days
— you own a workstream end to end on a live deployment: your data pipelines run in production, your weekly client steering happens without your lead in the room, and you've extracted at least one dataset that existed only in someone's head.
At 12 months
— you carry a deployment scope autonomously, the client's finance function has validated gains your work contributed to, and you've fed at least three field-driven improvements back into a partner product's roadmap. The strong version of you is mentoring the next cohort.
Your profile
Junior, hungry, high-slope. 0–3 years of experience. We care about what you can become in 18 months, and we've designed the training to get you there.
Must-haves
(non-negotiable)
- You code, for real. Python or equivalent, comfort with SQL and APIs. You have built working things — side projects, internships, school projects that shipped. We will test this. You don't need to be brilliant yet; you need to be solid and fast-learning.
- Structured problem decomposition. Given a messy, ambiguous problem, you can break it into parts, sequence them, and start with something that works. This is the single most tested skill in our hiring process.
- You want the field, not the office. Factories are loud, data is dirty, stakeholders are skeptical. That description excites you rather than scares you.
- You hold a conversation with a skeptical adult. You listen more than you talk, you ask questions that show you did your homework, and you can explain something technical to someone who isn't.
- Professional French and English. Client floors run in French; the company and its tooling run in English.
Nice-to-haves
(we'll be glad, we won't require)
- Exposure to industrial environments: an internship in a plant, a supply chain project, a family business in manufacturing, an apprenticeship (alternance) in operations.
- A quantitative backbone: engineering school, applied math, physics, or an equivalent you can demonstrate.
- Frontend basics (React/TypeScript) — a large share of real FDE work is building small apps clients actually use.
- Any evidence of ownership: a club you ran, a product you launched, a system you kept alive.
We'll teach you
(this is what the 3-week bootcamp and embedded apprenticeship are for)
- How a factory actually works, end to end, and how to read its P&L.
- Our deployment method: problem framing, data archaeology, target state design, MBB-grade project management.
- The products you'll deploy, the tooling, and the AI-native way we build.
Profile shapes that fit
- Fresh graduate of an engineering program with real code and at least one industrial or operations exposure (Arts et Métiers, INSA, ICAM, UTC, Centrale, Mines, X, or equivalent — the school matters less than the evidence).
- Junior developer or data engineer (1–3 years) who is bored of tickets and wants ownership and client contact.
- Junior consultant who discovered they'd rather build the system than write the recommendation.
- Apprentice/alternant from an industrial group who codes on the side and wants both worlds.
This role is probably not for you if…
- You want a pure engineering job with clean specs and no client in the room. Here the client is the room.
- You need certainty about what next month looks like. Deployments are assigned by need, and sites change.
- You're more comfortable describing a solution than shipping one. We measure output in running systems.
- Travel is a dealbreaker. Expect meaningful time on industrial sites, mostly in France, sometimes wider Europe.
- You go quiet when things get hard. We surface problems early, loudly, with context.
Career path
This is designed as an accelerated track, not a staffing pool. FDE → Senior FDE (owns accounts) → Lead FDE (owns a portfolio of deployments and a team) is the spine. Beyond it, two destinations inside the Devoteam group: the delivery leadership of The FDE Company itself as it scales across Europe, or an entry into the Devoteam partnership — open, through the group's partner process, to those who prove they can both bring and deliver recurring business at partner scale. Scope here is earned by what your deployments do for clients' P&L, not by tenure.
Working conditions
Office / Remote:
Paris base; expect 40–60% of time at client sites during active deployments
Travel:
France primarily; Europe as we scale
Training:
3-week paid intensive bootcamp at start, continuous training after
Equipment:
Full AI-native tooling stack — you'll work the way the best 2026 engineers work
Compensation:
€52–62K base + up to €8K deployment bonus on validated client gains
Our recruitment process
- Sourcing + factual screen
- Background deep dive & soft skills
- Technical screen
- Founder fit
- Values
- References Check
A word to our candidates
Every generation of engineers gets one role that compounds faster than the others. In 2010 it was mobile. In 2015 it was data. Right now it is this one: the engineer who can make AI systems produce real value inside real operations. The market is desperate for this profile, almost nobody trains it, and we built a company specifically to train it. You'll work harder than your friends in their first job. You'll also learn a decade of business, engineering, and human judgment in about two years, with the P&L receipts to prove it. If that trade sounds right, we want to hear from you. We'll be honest with you the whole way through — we'd ask the same back.
This is a saved copy of a posting published elsewhere. Postings get taken down without notice — check the employer's site before applying. mentors.coach is not the hiring party.