Consulting and tech lead assignments
An engineer with experience in defence, energy and space. I designed a complete information system, from architecture to operations.
Experience in large groups
Defence, energy, space, aeronautics, AI research.
Organizations listed as experience references. Logos remain the property of their owners.
Skills
Security, quality, operations: the foundation of a senior consultant.
Stack
Know-how: Idempotency, concurrency locks, workers, scheduled jobs, workflows and state machines.
Stack
Know-how: PDF generation, electronic signature, field photo capture, technical SEO.
Stack
Know-how: HttpOnly sessions and CSRF protection, AES-256-GCM encryption, rate limiting, server hardening.
Stack
Know-how: Reproducible deployments, rollback by tag, monitoring, alerting and deduplicated pull-mode backups.
Stack
Know-how: Online payment, ERP, electronic invoicing, calendars, transactional emails, and point-of-sale hardware.
Stack
Know-how: Engineering degree in AI and robotics: ML, deep learning, reinforcement learning, NLP, explainability.
Artificial intelligence
The point is not to use AI, but to industrialize it without losing control.
Engineer first. AI executes; I design, decide, review and prove.
I do not delegate my understanding: every change is reviewed, and I keep the final say.
Versioned specs split into deliverables with acceptance criteria; one instruction contract per repository.
Evidence: A 28-deliverable roadmap drove the SSO rollout across 9 applications.
A plan approved before any code, specialized sub-agents, worktrees to parallelize without conflict.
Evidence: A fresh session executes another one's plan: no confirmation bias.
Deterministic hooks, restricted permissions, no secret handed to AI, explicit human escalation.
Evidence: Guardrails are technical, not just instructions.
Tests written first, then multi-agent testing in a real browser; every verdict is re-measured.
Evidence: 5 campaigns, over 800 cases, up to 11 agents in parallel.
Every change reviewed, architecture and security calls made by me. No code I do not understand gets merged.
Evidence: I remain accountable for what goes to production.
Reusable skills, a wiki updated in the same patch, memory shared between agents.
Evidence: Active watch: spec-driven development, MCP, Claude Code, Codex.
Shared rules rather than individual habits.
An instruction contract per repository, versioned specs and a common definition of done.
Reusable workflows (feature, acceptance testing, diagnostics) versioned with the code.
Hooks, permissions and mandatory reviews: a technical frame, not just a declared one.
Helping developers move from improvised prompting to a measurable engineering practice.
Case studies
For each project, the architecture decisions and the engineering challenges.
Replacing a legacy tool with an in-house platform, built from scratch.
An internal web application for cataloguing and search, built from scratch by a team of seven whose architecture and code review I own.
13 interconnected apps: sales, workshop, rentals, invoicing and e-commerce
A complete information system for an SME: central API, e-commerce, ERP and point of sale, SSO and 6 business apps.
E-commerce API, storefront and back-office, with an ERP-driven catalogue
A 3-app headless e-commerce platform: ERP-driven catalogue, Mollie payments, legal invoicing and returns handling.
One identity, a second factor and central revocation across 9 applications
Self-hosted OIDC single sign-on for 9 applications: passkey or TOTP MFA, server-side tokens and migration with no lost access.
Dolibarr 22 with 8 custom modules, a connected POS and e-invoicing
Dolibarr 22 ERP with 8 custom modules: hardware-connected point of sale, offline inventory and a 4× faster product list.
One backend for the workshop, rentals, sales, invoicing and bank reconciliation
Central Node.js API: ~250 endpoints, 44 data models, quotes signed online, idempotent invoicing and server-side OIDC SSO.
Engineer first: AI executes; I design, decide, review and prove.
Versioned specs, skills and multi-agent browser testing: an AI pipeline under engineering control, with over 800 test cases run.
2 hardened servers, zero exposed app ports, tag-based rollback and pull backups.
Two hardened, monitored production servers: no exposed application ports, tag-based rollback and three-tier backups.
On assignment
Ramp up fast, secure delivery and leave a team more autonomous than when I arrived.
Understand the existing system, the business and the constraints before proposing anything.
Reasoned decisions and a breakdown into verifiable deliverables.
CI/CD, automated tests and systematic code review.
Reproducible deployments, metrics, logs and alerts.
Documentation, runbooks and team upskilling.
Terms
Adapts to the assignment and to your organization.
Embedded in your teams on time and materials, or fixed price on a defined scope.
Remote, or hybrid around Toulouse depending on the assignment.
Start date and length agreed based on the assignment schedule.
Other needs
The same background, for another need.
Context, stack and timeline: I reply with my analysis and my availability.
