Articles
What each project taught me about automating operations with AI: how to map the manual process, where to keep a person in control and how to validate before paying for AI.
AI in operationsAutomationDiscoveryDesign systems
- 10AI in operations
The AI proposes, a human confirms
In a legal AI agent and in a photo-based ordering system, the rule was the same: everything the machine produces is a proposal until a person approves it. The work is deciding where that approval sits, what gets locked, and what shows up before the click.
- 09AI in operations
Reading documents in bulk with AI
The machine can read hundreds of pages. What decides whether you can trust it is what it does with the page it couldn't read, the sentence with no source, and the line it didn't recognize.
- 08Automation
Map the manual process before you put AI in it
Before automating anything, I write down who does the work, what comes in, what goes out, where it goes wrong, and what a mistake costs. A distributor's photo-based ordering shows why that order matters.
- 07Automation
Custom automation with AI, in weeks rather than months
Three systems built in 11 days, 2 weeks, and about 3 weeks, each one taking a manual task off someone’s plate. AI wrote most of the code. What shortened the timeline was the method around it.
- 06Discovery
The Double Diamond in practice for choosing what to automate
Almost every automation request arrives with the solution already attached. The first diamond exists to find out what the problem is, and it decides whether the AI will take manual work off someone's plate or just move it somewhere else.
- 05Design systems
A design system that AI agents follow
An interface rule is only followed if it's written where the people building read. I learned that on a design system for product teams and confirmed it when the ones building the screens became a team of 13 AI agents.
- 04Automation
Prioritizing what to automate, without guesswork
In Glint, the biggest gap against the market was a manual process: refunds requested by email. Prioritizing meant finding that work, ordering the rest by what depended on what, and writing down what would stay manual, on purpose.
- 03Automation
Validate the automation flow before paying for the AI
In a photo-based ordering system, a simulated reading let me validate the entire flow without paying for a single real one. What it didn't catch taught me more than what it confirmed.
- 02Discovery
From sticky note to the process worth automating
An inventory of manual tasks shouldn't end in a list of automations. It should end in a tree that connects the outcome the company wants to the work people do by hand.
- 01Automation
Interview the people doing the manual work before you automate it
To find out what to automate first, I don't ask people what they want from the system. I ask them to show me the last time they did the work, and then I listen.