AI Data Engineering Lead — Build Production AI Pipelines

Fabriq; Paris; Hybride; EUR 90 000 - 100 000; Plein temps

Avantages offerts par ce poste

7 weeks of paid leave per year

Alan health insurance

Swile meal card

Company event every quarter

Annual offsite

Résumé du poste

Fabriq in Paris is seeking a senior data engineer to join Bertrand’s team, working directly with our CEO and a dedicated Product Manager to ship AI features across the roadmap. You will manage 2 Senior AI Software Engineers with a third hire planned, and help build the data foundations for 700 sites.

You will design resilient ETL/ELT pipelines, lead data governance, and co-build a multi-site leadership tool, shaping the roadmap with the PM. Hybrid Paris-based setup.

Qualifications

• 5–8 years in data engineering or analytics, plus 1–2 years leading a team.
• Experience designing and running production ETL/ELT pipelines with governance.
• Contributed to AI features in production, data prep, evaluation, or integration.
• Hands-on daily work with SQL, Python, pipelines; comfortable with pair programming and TDD.
• Experience fine-tuning or deploying open-weight models is a plus.

Responsabilités

• Lead and coach a small team of data engineers while delivering features.
• Build data pipelines and features in TypeScript/AI stack.
• Design resilient ETL/ELT pipelines with governance for data foundations.
• Collaborate with PM to shape roadmap and deliver co-builds across sites.

Connaissances

Data engineering

Team management

SQL

Python

Pair programming

TDD

AI features in production

Outils

TypeScript

AWS Bedrock

Description du poste

Context: Lighthouse is becoming a full product team. You join Bertrand’s team (VP Engineering), working directly with our CEO and hand in hand with a dedicated Product Manager. You manage 2 Senior AI Software Engineer, with a third hire on the way, target of 3 to 4 engineers. The team is shielded from run work and focuses on the AI product.

The Goal: Ship the 3 AI features on the roadmap to production within 12 months, one per persona, on data foundations ready for our 700 sites.

Key Challenges:

• Player-coach leadership: You split your time between leading and building. You build pipelines and features yourself and pair with the team, while coaching, setting standards and hiring.

• Knock down the first dominoes: You take the Problem Solving Coach from POC to production, then launch data synthesis for managers and AI analytics for leadership. The agentic stack is built from scratch in TypeScript on Deno, with frontier models through AWS Bedrock.

• Data foundations: You design and build resilient, well-governed ETL/ELT pipelines on shop-floor data, so that each building block makes the next feature easier to build.

• Co-build and roadmap: You deliver our first co-build, which tailors our features and lays the groundwork for a multi-site steering tool for leadership. With your PM, you fully shape the roadmap: the foundations are solid and much is still to be decided.

Must-haves:Experience: 5 to 8 years in data engineering or data analytics, including at least 1 to 2 years managing a team (feedback, growth, hiring).Pipelines in production: You have designed and run data pipelines (ETL/ELT) in production, with real stakes around quality, resilience and governance.Data for AI: You have contributed to AI features in production, on data preparation, evaluation or integration into an existing product.Hands-on every day: You still build regularly (SQL, Python, pipelines), use AI every day in your work, and are comfortable with pair programming and TDD.Nice to have: These skills remain optional.Experience fine-tuning or deploying open-weight models in production.Some software development experience, ideally in TypeScript, the language of our stack.Compensation and setup :Package: €90K to €100K gross per year.Perks: 7 weeks of paid leave per year, Alan health insurance, Swile meal card, a company event every quarter and an annual offsite.Location: Our offices are in Paris 10th and we favor a Paris-based profile, with a hybrid setup.Start date: From January 2027.Languages: Fluent French and English.