École hôtelière de Lausanne
BSc, Hospitality
Elite training in reading people, de-escalating operational tension, and stakeholder communication — the soft infrastructure that decides whether software is adopted or abandoned on the floor.
The combined profile
Three worlds smaller companies rarely find in one partner. I bring them together.
MIT Schwarzman College of Computing
Data Science, Machine Learning & Data Engineering
Production machine learning, data engineering, generative AI and agent systems — enough breadth to choose the right technology and enough depth to make it operate beyond a demo.
The Code Venture · Founder & CSO
30+ Swiss startups · 13-person team · €1M in 3 years
Founded and scaled a consulting agency supporting Swiss startups from idea to launch. Built the hiring and development culture behind a highly skilled team with low turnover — then applied that founder discipline inside enterprise Data & AI leadership.
Verified learning
Source documents available
Formal depth behind the applied work.
Selected credentials spanning data science, AI development, local language models and generative-AI instruction.
MIT IDSS / SchwarzmanData Science & Machine Learning: Making Data-Driven DecisionsCompleted January 2023
MindStudioAI Developer Certification ProgramCompleted 2024
Pragmatic AI Labs / edXApplied Local Large Language ModelsIssued December 2025
GSDCGenerative AI Foundations · Training ProviderGlobal Skill Development Council
Accountability map
Philippe leads. Specialists strengthen the build behind him.
You hire one accountable lead, not a loose bench. Philippe owns the business problem and the product direction; focused specialists support his delivery where their depth is needed.
Accountable lead
Founder · Forward-Deployed AI Lead
Philippe owns diagnosis, product strategy, domain translation, architecture and the client relationship. He remains at the head of the engagement from the first bottleneck through adoption.
StrategyDomainArchitectureAdoption
DS
David Santos
AI/ML engineering
Supports production engineering, model development and tooling when the product requires deeper build capacity.
BR
Bruno Reis
Security & data integrity
Supports security and privacy decisions where operational or client data needs additional scrutiny.
Selected specialists
Domain · integration · delivery
Additional depth is selected for the exact operating problem and works under Philippe's direction.
You do not manage the bench. Philippe does.
2026 engineering portfolio
The depth normally reserved for enterprise AI programmes. Applied to focused business problems.
These systems show the range behind the product catalogue: governed agent architecture, sovereign deployment, compliance workflows and retrieval over operational evidence.
Flagship research build · June 2026
Datakult 2.0: governed AI self-improvement
Philippe built and operated a cognitive software stack in which stochastic LLM workers can propose repairs, while deterministic gates and an append-only audit record retain authority. A 60-day technical study documents the operating pattern, its evidence and its limits.
60-day operating study42-page paperDeterministic promotion gates
Read the technical paper
Multi-agent compliance review
A governed agent workflow for a Swiss private-wealth environment, dividing document extraction, rule analysis and audit support into specialised responsibilities.
Air-gapped LLM operations
A sovereign deployment architecture for a Llama-class model on Red Hat OpenShift and Kubernetes, designed for controlled infrastructure and strict data boundaries.
Operational safety retrieval
A retrieval system connecting more than a decade of safety reports to a large regulatory-control library, helping teams surface relevant precedent from field evidence.
Client identities and implementation detail are shared only where permissions and confidentiality allow.
Track record
A decade productionising AI inside regulated enterprises.
A generalist who combines founder agility with enterprise rigor — building a services business from zero, then deploying production AI for regulated operations and national infrastructure.
{{ t.period }}
{{ t.role }} · {{ t.org }}
{{ t.desc }}
Languages
English · Native / BilingualPortuguese · Native / BilingualSpanish · ProfessionalFrench & German · Working
Global mobility
Dual Swiss & Portuguese national, 100% prepared to deploy on-site globally — UAE/Dubai, US, EU — for 6+ months to embed with strategic customer teams.
Community builds
Open Brain — a retro terminal-style sandbox for mastering multi-agent orchestration. ETH Juniors Alliance — a Horizon 2030 (€1.5M) consortium for an on-chain AI learning metaverse.
Operating principles
How we work inside client operations.
⛓
One partner, not disconnected tools
One partner who connects solutions across the value chain and keeps them running.
⌥
Workflow-first implementation
We map the real workflow, not the org-chart version. Solutions go where teams already make daily decisions.
⚇
Side-by-side delivery
We work inside operations teams so adoption happens in context, not in a slide-deck handoff.
♺
Continuity by design
Transfer of ownership is built in, so your team runs the system without permanent vendor dependency.
Talk to the engineer who'll be in the room.
Bring the operational problem. Philippe will help establish whether a focused product has a credible first use case.