Expertise
From AI adoption to the foundations that carry it
Five skills, ordered by what organisations ask me for first. The first two make up the AI Adoption Lead role; the next three are what make it possible.
The role I offer
AI Adoption Lead
Taking AI from isolated trials to measured use: choosing use cases, setting data and access boundaries, defining how results are evaluated, equipping and training teams, then tracking adoption over time.
- Measurable pilots, with acceptance criteria
- Guardrails: data, access, traceability
- Training and a network of internal champions
A role I offer on engagements, backed by the projects in my case studies.
Operational AI
LLMs, agents, RAG, OCR and vision models, open-weight or self-hosted models: used where they bring measurable value, with clear data boundaries and human validation.
The symptom
When AI was tested without a criterion or a frame, and never left the pilot.
What I deliver
- Use cases chosen against a written success criterion
- Comparative benchmarks before any model or vendor choice
- Bounded assistants and agents, with human validation
- Private or local hosting when the data requires it
Typical tools
- LLM
- Agents
- RAG
- OCR and VLM
- Open-weight models
- Evaluation
Related proof
Automating after-sales returns to manufacturers, with a trace of every request
Santécie’s after-sales return files are routed to the right manufacturer forms and filled automatically, with an audit trail for every request.
Training developers in agentic engineering, in eight sessions
I am preparing an eight-session path, designed to take developers from occasional use of an assistant to verifiable agentic engineering, including knowing when to do without it; first session on 16 October 2026.
AI adoption and training
Taking AI from isolated trials to a team practice: usage rules, guardrails, agentic engineering training for developers, internal champions and usage measurement.
The symptom
When everyone tinkers with their own assistant, with no shared rule or measurement.
What I deliver
- Usage frame: allowed data, access, traceability
- Agentic engineering programme for developers, on your repositories
- Team primitives: AGENTS.md, skills, verified workflows
- Adoption indicators defined before rollout
Typical tools
- Workshops
- Hands-on labs
- AGENTS.md
- Skills
- Usage measurement
Data and automation
Connect scattered data, automate what should be automated, and keep a trace of every decision the system makes.
The symptom
When the information exists but never reaches the right place.
What I deliver
- Map of flows and sources
- Routing, processing queues and automations
- Run traces and replay without duplicates
- Impact indicators defined before they are measured
Typical tools
- SQL
Python
PostgreSQL- BullMQ
- Windmill
- APIs
Custom business applications
Internal tools shaped around your real processes: case handling, forms and approvals, back offices, mobile apps, admin portals.
The symptom
When spreadsheets, email and re-keying stand in for a system.
What I deliver
- Process scoping and MVP definition
- Journeys designed with field teams
- Full-stack development, web and mobile
- Production release, documentation and handover
Typical tools
TypeScript- Vue.js
NestJS
Prisma- React Native
- PostgreSQL
Related proof
Turning a training method into a mobile app and its back office
The NFC method exists as a product: a mobile app for members and a back office for the team, on a single data foundation.
A paid-leave request application for CICP2R
On the employer side, paid-leave requests and days to pay are entered in a dedicated application, then go through an explicit validation.
Architecture, environments and hosting
Foundations that last: architecture, reproducible developer environments, deployment, maintenance and documentation, with security considered from the start.
The symptom
When technical debt and hosting slow down every change.
What I deliver
- Architecture audit and recorded decisions
- Developer environments described as code
- Integration with your existing information system
- Deployment, monitoring and maintenance
Typical tools
Docker- Terraform
Coder
Railway
Cloudflare- CI/CD
Related proof
Zedflow, agent flows you can read, compose and control
With Zedflow, an agent flow is read, validated and composed like a program, and sensitive decisions wait for a human approval.
Choosing a health-card reading engine on evidence, not on a demo
The choice of a reading engine for supplementary health insurance cards rests on a reproducible benchmark, not on a sales demo.
Let’s talk about what AI could change for you
A 30-minute discovery call, by video and with no commitment, to talk about your project, your constraints and the best next step.