Mathieu Bernard
CTO · Technical leadership, audits, project rescue, AI in production
contact@mathieubernard.tech · linkedin.com/in/mathieu-bernard-cto · mathieubernard.tech
CTO since 2013. I built and ran three development centres, hired and onboarded more than 400 people, and today I design AI agents that run in production. I also work as a fractional CTO: audits, due diligence, project rescue, team leadership.
Experience
2025 — present
CTO and partner · Etixio
- Building and leading dedicated, distributed development teams in Madagascar and Mauritius
- Technical audits and due diligence before a fundraise, sale or acquisition
- Fractional technical leadership, steering teams and vendors, hiring CTOs
- Rescuing projects that drift
- Designing AI agents in production: an AI employee platform (Symfony, React), a base for custom agents (TypeScript), AI Gateway, an open-source LLM gateway
2013 — 2025
Group CTO, partner, then CEO · Bocasay
- Created and ran three development centres, in Madagascar, Vietnam and Mauritius
- 170 developers in production for French and international clients
- More than 400 people hired, trained and placed in teams
- Technical strategy, delivery quality, budget and profitability; PHP / Symfony, JavaScript and mobile stacks
2010 — 2013
Developer, then project manager · Didhaptic · Netapsys
- PHP and C++ for a virtual-reality dental surgery simulator
- Leading Symfony projects
Skills
- Technical leadership
- Roadmap, budget, hiring, vendor management, handover
- Audits
- Code, architecture, security, technical debt estimated in days, due diligence
- Development
- PHP 8, Symfony, TypeScript, Node, React, PostgreSQL, Docker, FrankenPHP
- AI
- Agents and tool calling, MCP, local models (llama.cpp), evaluation, costs, guardrails
- Practices
- Claude Code, automated tests, CI/CD, code review, documentation for agents
Training I deliver
- Steering tech and AI without being technical · 2 half-days
- Claude Code for development teams · 4 half-days
- Building an AI agent, from idea to production · 6 half-days
- From AI prototype to production · 4 half-days