Back to portfolio
Freelance services

Senior data and AI delivery for teams that need production-grade results

I step in when the architecture matters, the execution window is tight, and the solution has to survive beyond the demo. Data platforms, AI products, and end-to-end delivery from framing to rollout.

Open to new engagementsParis, France · Remote across Europe8 years across data, product, and architecture

Where I step in

Four service lanes, one delivery standard

The mandate changes, but the bar stays the same: production-ready architecture, clear ownership, and no decorative complexity.

Data foundations

Data platform architecture

I design or restructure the data layer when the warehouse exists but the operating model is still blurry: brittle pipelines, weak modeling standards, unclear ownership, or runaway costs.

Best when

You need a platform shape that analytics, engineering, and business teams can all work with.

Target architecture and key technical decisions
Warehouse, ingestion, and transformation design
Quality, ownership, and cost guardrails
Migration or scale-up plan tied to real workloads

Recent context

Pernod Ricard, Air Liquide, DataPulse

Typical stack

SnowflakedbtMatillionPostgreSQL
AI product delivery

Production AI engineering

I take AI features past the impressive demo phase: grounded retrieval, observable orchestration, evaluation loops, streaming UX, and operational controls that hold once usage grows.

Best when

You have a real user workflow in mind and cannot afford a system that sounds smart but fails under load.

RAG and agent workflow design
Tracing, evaluation, and cost monitoring
Model routing, guardrails, and fallback logic
API and product integration for real users

Recent context

Comply-Agent, Spectre, Agentic Trading Hub

Typical stack

LangChainQdrantLangFuseFastAPI
End-to-end builds

Full-stack product delivery

I ship the product surface around the architecture: frontend, APIs, auth, billing, observability, and deployment. Useful when the team needs one person who can carry both product momentum and technical coherence.

Best when

You want a senior builder who can keep the whole product moving instead of handing problems across four specialists.

Product architecture across frontend and backend
Authentication, billing, and core user journeys
Deployment workflows and production monitoring
A maintainable handover path for the next team

Recent context

CoffreZero, PrettiOps, Cartely, DataPulse

Typical stack

Next.jsTypeScriptFastAPIStripe
Senior leverage

Technical advisory and fractional leadership

I step in when the team needs senior judgment before scaling the org: architecture reviews, roadmap shaping, delivery governance, and technical coaching with direct execution behind the advice.

Best when

You need sharper decisions now, but not necessarily a full-time executive hire yet.

Architecture reviews and decision records
Roadmap shaping with delivery tradeoffs made explicit
Technical coaching, review cadence, and unblock plans
Stakeholder alignment around scope, risk, and sequencing

Recent context

Enterprise transformation and founder-led builds

Typical stack

ADRsRoadmapsReviewsExecution governance

Typical situations

A proof of concept needs to become dependable software

The idea is validated, but no one trusts the path to production yet.

The stack works, but the operating model is fraying

Metrics drift, ownership is fuzzy, and delivery slows every quarter.

You need senior execution before building a full team

There is real momentum, but not enough senior technical shape around it.

Engagement models

01

Scoped build

A defined problem, clear deliverables, and a short path to something usable.

Usually 4 to 12 weeks

02

Embedded freelance mission

Inside your team as a senior architect or product engineer, full-time or part-time.

Usually 3 to 12 months

03

Fractional technical lead

For teams that need senior direction, reviews, and sequencing without a full-time leadership hire.

Ongoing cadence

Have something complex to unblock?

Send the context, the constraints, and where the team is stuck. I will come back with a delivery angle, not a generic pitch.