AI · AUDIT · DEVSECOPS · SOVEREIGNTY

Move AI forward.
Stay in control.

Since 1998, I have designed and modernised software platforms. My current work also involves LLMs, local AI, technical audits and team mentoring, with particular attention to security, GDPR, resilience and digital sovereignty.

Portrait of David Breton
David BretonSoftware architecture · AI · DevSecOps
Since 1998Engineering experienceSince January 2022Parallel consulting & auditAI · Security · GDPRPractical risk management
Rodin’s Thinker reimagined with illuminated neural circuits, symbolising a critical approach to AITHE THINKER · REIMAGINED
01/03
ARTIFICIAL INTELLIGENCECapability &
judgement.
AI · AUDIT · SECURITY · CAREER↓ SCROLL
HOW I WORK

Three areas of intervention.
Practical outcomes.

Engagements shaped around existing systems, teams and constraints.

Approach & deliverables ↗
01 / ARCHITECTURE

Assess systems and options.

Review architecture, integrations, dependencies and operational constraints. Identify risks and trade-offs.

Possible outputs Architecture map, findings, priorities and modernisation roadmap.

Read case studies ↗
02 / DEVSECOPS

Improve delivery and reliability.

Review pipelines, quality gates, containers, monitoring and recovery procedures.

Possible outputs Action plan, pipeline examples, release and recovery acceptance criteria.

See the approach ↗
03 / AI

Test use cases and their limits.

Compare local and cloud models, data constraints and team adoption approaches.

Possible outputs Feasibility assessment, technical comparison and workshops.

Visit the laboratory ↗

Have a specific challenge? Start with the scope and constraints.

Discuss your context ↗
CURRENT PRIORITIES

AI keeps moving.
Responsibility must follow.

AI integration, infrastructure audits, team coaching, security and technological independence belong in the same discussion.

01 / AI & USE CASES

Experiment. Evaluate. Share.

Containerised language models, generative tooling and workshops: testing practical usefulness without overlooking constraints and limits.

LLMChatGPTCopilotDocker
AI & team adoption
02 / SINCE JANUARY 2022

Audit and support.

Architecture assessment, secure delivery, CI/CD, monitoring and engineering practices, alongside long-term client assignments.

AuditDevSecOpsObservability
Areas of intervention
03 / DIGITAL TRUST

Security, GDPR, resilience, sovereignty.

Protect data, anticipate disruption, verify restoration and preserve the ability to change providers.

DataReversibilityContinuity
Architecture principles
PRACTICE & METHOD

Explore AI.
Assess the existing.

Model inference, speech interfaces, automated delivery and recoverability — with a disciplined approach to technical assessment and decision-making.

01 / AI LAB

Models, voice and local AI.

Documented explorations of Mistral, NLLB, Whisper, Gemma and voice synthesis.

Explore the lab
02 / CONSULTING SINCE 2022

Audits and DevSecOps coaching.

Architecture, secure delivery, supervision and practical skill transfer.

See the approach
03 / DIGITAL TRUST

Security, GDPR and sovereignty.

Review infrastructure choices, vendor dependencies, access controls and recovery plans.

Compare options
PROFESSIONAL EXPERIENCE

Organisations. Challenges.
Practical delivery.

Hospitality, media, e-commerce, industry, banking and insurance: from software design to industrialised application delivery.

1998Career start
13Organisations
16Assignment periods
1998Foundations
2001Eurosport · Media
2009Performance
2016SaaS & Quality
2019DevOps
2024Platforms
Explore the career timeline Download English CV
SELECTED PROJECTS

Architecture decisions.
Documented work.

Selected technical assignments and the engineering choices involved.

All case studies ↗
CASE STUDY

Modernising CI/CD and API delivery

Transition from Jenkins to Azure DevOps and API platform modernisation with MuleSoft 4.

Git/YAMLAzure DevOpsSonarQube / ReadyAPI
Explore this project ↗
CASE STUDY

Publishing and monetising sports content

Content platforms for Eurosport and partners including TF1, across several markets.

Content sourcesBizTalk / MSMQEDI / EAI
Explore this project ↗
CASE STUDY

Standardising heterogeneous applications

Containerisation of PHP, .NET Core, Ruby and Python applications with repeatable CI/CD.

ApplicationsDockerGitLab CI/CD
Explore this project ↗
CASE STUDY

Improving SaaS engineering and quality

Architecture work supporting software services and delivery reliability.

CodeTFSSonarQube
Explore this project ↗
CASE STUDY

Engineering for high-volume e-commerce

Application development and performance optimisation in an online retail environment.

ASP.NETServicesSQL Server
Explore this project ↗
ENGINEERING EXPERTISE

Architecture and security.
AI and practical delivery.

From application architecture to secure deployment, connecting use cases with operational constraints.

All areas of expertise ↗
01

Architecture

Distributed services, APIs, SOA, DDD/CQRS, .NET and data integration.

02

CI/CD & platforms

Azure DevOps, GitLab, Jenkins, Git, containers and OpenShift.

03

Quality & security

Code review, SonarQube, Trivy, DevSecOps, risk prioritisation.

04

Monitoring & recovery

Prometheus, Thanos, Grafana, ELK, alerting, backups and restore drills.

05

AI engineering

LLM experimentation, integration, inference infrastructure and data boundaries.

06

Mentoring

Clean Code, TDD, Docker, agile delivery and operational adoption.

THE COMMON THREAD

Good technology is not measured by capabilities alone. What matters is whether it can be understood, maintained and passed on.

ENGINEERING PRACTICE

Real constraints.
Thoughtful trade-offs.

AI, security, recoverability and automation are engineering questions with human consequences.

Explore experiments ↗
EXPERIMENTS01

Containerised LLMs and translation

Mistral 7B and NLLB in Docker: model packaging, inference boundaries and data-location trade-offs.

EXPERIMENTS02

Voice assistant architecture

Wake word, Whisper/faster-whisper speech recognition, Gemma, OpenVoice/MeloTTS/Piper. Latency, permissions and local processing are central challenges.

EXPERIMENTS03

Local GPU, dedicated cloud or API?

Evaluation of memory, availability, operational cost, data sensitivity and provider exit options.

Cover of Therefore IA-M? — Je pense, donc je douteCE QUI NOUS DÉPASSE 01
BOOKS & IDEASDisponible sur Amazon

A different way
to explore AI.

CE QUI NOUS DÉPASSE · FIRST BOOK

Therefore IA-M?

Je pense, donc je doute

A personal essay about artificial intelligence, its potential, the future of work and the responsibility to question our assumptions.

ABOUT

Technology.
Several perspectives.

Nearly thirty years in software engineering offer perspective, not certainty about what comes next.

Software architecture remains central to my work. Since January 2022, I have complemented client assignments with technical audits, mentoring and consulting. AI now features prominently, with particular attention to privacy, security, resilience and data control.