AI Dev Setup
The essentials for continuing to develop with AI within a clear framework.
Discuss my project- AGENTS.md and project conventions
- Scripts, reproducible environment and agent workflow
- Targeted tests, lint and typecheck
AI Engineering Infrastructure
Keep building with AI.
A structured environment for your agents to evolve your product with clear rules, tests and validation.
Discuss my project
Your product evolves. Its context needs to keep up.
You built your application with AI. Each new prompt or agent means explaining the project again, and a fix sometimes breaks another user journey.
I turn your AI-built or vibe-coded project into a structured development environment where Claude, Codex, Cursor and your future agents can work effectively.
The core deliverable: context that agents can use, kept in your project and available to your team.
From isolated prompts to verified changes.
Every change follows a known procedure. Business rules and checks stay with the code all the way to production.
- Specification
- Agent
- Conventions and tools
- Tests
- Validation
- CI
- Production
A system shaped around your project.
We choose the deliverables based on your architecture, critical journeys and the way you already use AI.
Project rules and context
AGENTS.md, architecture, conventions, business rules, APIs and data schema. Sensitive areas, migration procedures and acceptance criteria are documented.
Agent harness
Standard scripts and commands, a reproducible environment, targeted tests, lint and typecheck. Your agents know how to start, change and verify the project.
Specialised skills
Procedures tailored to your code: create-feature, database-migration, debug-production, deploy or write-tests, with the context they need.
Guardrails and delivery
CI/CD, validation before commits and PRs, controlled access, permissions, sandboxing and secret management within the agreed scope.
Observability and feedback
Logs, errors and monitoring to understand incidents. Lessons feed back into documentation, tests and procedures.
Useful connections
MCP for GitHub/GitLab, documentation, databases, observability or ticketing when access offers a concrete benefit. RAG when your business knowledge justifies it.
Three levels, with a scope we agree together.
Describe your project through the contact form. Together, we will define the deliverables, scope and timing that fit your needs.
Recommended foundation
AI Engineering Infrastructure
The foundation for developing, delivering and operating your product.
Discuss my project- Everything in AI Dev Setup
- CI/CD and specialised skills
- Architecture documentation and business rules
- Observability, permissions and secret management
Agentic Engineering System
Advanced workflows for projects that need several agents and tools.
Discuss my project- Everything in AI Engineering Infrastructure
- Specialised agents and advanced workflows
- MCP according to access needs
- RAG when knowledge volume and quality justify it
Unblock, structure, support.
Rescue / Production Readiness
An essential journey is blocking your customers? The unblock sprint fixes, tests and deploys that journey before we structure what comes next.
Explore the unblock sprintAI Engineering Infrastructure
Your product works: I set up the context, tools and validation to keep developing it with AI.
Engineering Partner
Monthly support quoted to scope: architecture, reviews of AI-generated changes, infrastructure, production and incidents.
Define the next development framework for your product.
Describe your application, stack, AI tools and current difficulties. I propose a level and scope that fit your project.
Discuss my project