Small business AI consulting

Skip the AI hype. I help small businesses adopt AI where it actually moves revenue — building your own agents, configuring Claude Code and opencode, setting guardrails, and locking down basic AI security so real data stays safe.

In short: most small businesses don't need a moonshot AI project — they need two or three boring automations that save hours a week, an agent CLI set up properly, and guardrails so their data stays safe. That's what I deliver, with fixed-fee scopes and no jargon.

Four ways I help small teams

Build your own AI agents

Purpose-built agents for the work you actually do — drafting, triage, research, summarising — built with the guardrails that keep them predictable and useful.

Claude Code / opencode setup

Installing and configuring an agent CLI for your team — models, permissions, context files, and the workflow habits that make it a force multiplier instead of a novelty.

Guardrails & workflow design

The prompts, policies and safety rails that keep AI output on-brand and on-track — including when to use a local model (Ollama) instead of the cloud.

Basic AI security review

A plain-English pass over where your data goes: which tools train on it, where secrets could leak, and the simple rules that keep you out of trouble.

Not a business? Try personal AI coaching.

Same approach, one-on-one — learn to actually work with Claude: how to talk to it, what skills are, what MCP is, and how to get it doing the work instead of just answering questions.

Explore personal AI coaching

Built by someone who runs this stuff in production.

I build and run air-gapped, local AI pipelines — ComfyUI, FLUX, WAN, LTX, Qwen3-TTS — on real hardware for enterprise clients under NDA. That means the advice you get for your small business is grounded in systems that already work, not vendor slide decks.

Frequently asked questions

Where should a small business start with AI?

Start with the boring, high-ROI tasks: drafting, summarising, data entry, customer replies and internal search. I help you find the two or three automations that actually save hours per week — and skip the ones that only look impressive.

Do I need my own AI agents?

Maybe. Some businesses get most of the value from configuring a good agent CLI like Claude Code or opencode and writing better prompts. Others genuinely benefit from a purpose-built agent. We decide based on your workflow, not on hype.

Is it safe to put my business data near an LLM?

It can be, with guardrails: knowing which tools train on your data, keeping sensitive data local where possible, using a local model like Ollama for private work, and never pasting secrets. I set those guardrails up with you.

What does a small-business AI engagement cost?

Engagements are scoped and fixed-fee where possible: a short audit, a setup-and-training session, or an ongoing advisory retainer. You will always know the price and the deliverables before we start.