// services/ai-assisted-engineering
AI-assisted engineering with Claude Code
AI coding tools are only as good as the context they get. Out of the box they write code that works but ignores your module boundaries and conventions, and someone has to rewrite it.
I set teams up with Claude Code the way it works in production: skills and rules written for your codebase, MCP integrations with the tools you already use, and agent workflows with review built in. I also build LLM features into products on the Claude API.
What's included
- Codebase-specific skills and rules
- MCP integrations such as GitHub, Jira, Figma and Slack
- Agent workflows with programmer, reviewer and QA subagents
- Parallel agent sessions, each isolated in its own git worktree
- LLM features on the Claude API: tool use, prompt caching, streaming and evals
A good fit when
- Your team uses AI tools, but the output doesn't fit your codebase
- You want AI to speed up delivery without lowering the review bar
- You want an AI feature in your product that is reliable enough to ship
// proof
Where I've done this
- Fintech · Wealth managementWhiteboxIntroduced Claude Code with codebase-specific skills, rules and MCP integrations, so AI code follows the app's module boundaries.Read the case study →
- Marketplace · AI engineeringPeer-to-peer car rental marketplaceAn agent-driven workflow that merged 400+ pull requests in four weeks, and an in-app support agent on the Claude API.Read the case study →
// how it works
A simple process
- 01 · discover
A free 30-minute call about your product, team and goals.
- 02 · propose
A short written proposal with scope, approach, timeline and price.
- 03 · build
Regular demos and transparent progress. You own all the code.
- 04 · handover
Documentation and knowledge transfer, so your team can carry it on.
Let's talk about your project.
A free 30-minute call, no commitment. I reply within one working day.