AI features & workflow automation
Most businesses don't need an “AI strategy” — they need specific hours back. The AI work I ship is practical: assistants that answer from your own data, vision that reads documents and photos into structured records, content generation tuned to your voice, and automations that quietly handle the repetitive work between your tools.
I've built production AI features ranging from a marketing engine that writes daily, context-aware promos for local businesses to vision that turns a photo of a water test strip into readings and prioritized next steps.
The common thread is guardrails: metered and cost-capped AI calls, human review where it matters, and an honest assessment of whether AI is the right tool for the job at all.
- AI features & assistants
- Vision & document AI
- Workflow automation
Related work
Bloom
An AI marketing engine for local businesses that writes ready-to-post promos every morning — tuned to the weather, nearby events, and each shop’s own slow days — then delivers them across Instagram, email, and SMS.
Salt Water
A water-chemistry tracker that reads a test strip straight from a phone photo using AI vision, then turns the readings into clear, prioritized maintenance steps.
Common questions
- Which AI models do you use?
- Primarily Anthropic's Claude, with the flexibility to use whichever model best fits the task and budget. Calls are metered and cost-guarded so bills stay predictable.
- Is my data used to train models?
- No. The APIs I build on don't train on your data, and your data stays in your own systems.
- Where does AI actually help?
- Anywhere someone repeats a judgment-light task on a screen: drafting, tagging, extracting, summarizing, routing. Discovery starts by finding those hours in your workflow.
Ready to talk about ai & automation?
Discovery is free: we talk through your goals and budget, and you get a clear scope and plan — whether or not we work together.