Solutions engineer · AI agents · Automation

I find what’s broken in a customer’s workflow and build the AI that fixes it.

I’ve built more than 20 voice agent demos for a voice AI platform, and two of the prospects who saw them became customers. Before that, I spent five years keeping enterprise systems running for companies in energy, FMCG, chemicals and petroleum.

Customer problems I’ve worked on.

A CRM for a driving school that ran on paper

Prototype in two days. Still in use.

1,063 students managed

Read the story

Products I made for myself.

Live on the App Store

Proof

An iOS app for people who keep going back to check the door or the stove. Each item on your exit checklist is ticked off with a photo stamped with the time and place, so you can look at the proof instead of driving home.

App Store

Open source

SnapAI

A native macOS app I added to the open-source SnapAI image-generation CLI by Code with Beto. I use it to make the logos for my other projects.

GitHub

Open source

Prompt-Lab

A local tool for finding where a prompt breaks before your users do.

GitHub

The same four steps, whatever the tool.

  1. 01

    Find what’s actually broken

    It’s rarely “we need AI”. It’s enquiries going cold, or someone copying data between two tools all afternoon.

  2. 02

    Put something working in front of them fast

    Two days, not two months. People tell you far more about a prototype than about a spec.

  3. 03

    Connect it to what they already use

    Salesforce, WhatsApp, the spreadsheet nobody wants to give up. Adoption depends on it.

  4. 04

    Keep going after handover

    The first version is a guess. Real users show you what the second one needs.

Hiring for solutions or deployment roles?

I’m looking for a solutions engineer, forward deployed or deployment role at an AI company, somewhere I’m close to customers and still building. Email is the quickest way to reach me.