Which AI providers do you work with?

Primarily OpenAI and Anthropic APIs, chosen per use case — the integration pattern matters more than the specific provider.

Can you add AI features to my existing product?

Yes — most AI engagements are adding a specific capability (an assistant, an automation, a RAG search) to an application that already exists.

Do you build AI agents that take real actions, not just chat?

Yes, with explicit tool definitions, permission boundaries and error handling around what the agent is allowed to do.

AI Engineering

AI development — assistants, agents and AI-powered product features.

I build AI-powered features that ship inside real products: assistants, agentic workflows, retrieval over private data, and automation connecting LLM APIs to the systems that already run your business.

LLM APIsTypeScriptNode.jsAutomationAPIs

Start a project See pricing

What I build

AI assistants and chat interfaces embedded in a product
Agentic workflows — multi-step, tool-calling automation, not a single prompt
Retrieval-augmented generation (RAG) pipelines over private or proprietary data
LLM API integrations (OpenAI, Anthropic and others) into existing applications
AI-powered automation connecting APIs, data sources and internal tools

Who it's for

Products adding an AI-powered feature to an existing application
Teams that want to automate a workflow with LLMs instead of manual process
Companies that need this built by an engineer who understands both the AI layer and the production system around it

How I approach it

The right tool, not AI for its own sake

Some problems don't need an LLM. The first question is always whether AI is actually the right solution, not the default one.

Reliable over flashy

Agentic and automated flows are built with explicit error handling, fallbacks and guardrails — not left to hope the model behaves.

Data and privacy-conscious

What data reaches which model, and under what handling, is a deliberate decision — not an afterthought.

Production-grade, not a prototype

AI features are built to the same standard as the rest of the product: typed, tested against real inputs, and monitored.

Frequently asked

Which AI providers do you work with?: Primarily OpenAI and Anthropic APIs, chosen per use case — the integration pattern matters more than the specific provider.
Can you add AI features to my existing product?: Yes — most AI engagements are adding a specific capability (an assistant, an automation, a RAG search) to an application that already exists.
Do you build AI agents that take real actions, not just chat?: Yes, with explicit tool definitions, permission boundaries and error handling around what the agent is allowed to do.