Skip to content
Pretesh Patel

Search posts

Let's discuss

The best thing about writing this site is the replies. If you are trying to get AI to stick inside a real organisation and something here landed, or badly missed, I would like to hear about it.

I am not selling or advising anything. I am just unreasonably interested in this problem, and I learn the most from people who are in the middle of it.

Things I never get bored of

A pilot that worked, then went nowhere

The demo landed, everyone was pleased, and six months later nothing has changed. This is the most common story in AI right now and the reasons are more interesting than people expect.

Adoption somewhere the stakes are high

Clinical settings, regulated industries, anywhere a person is liable for the output. The usual playbook assumes a tolerance for being wrong that these places do not have.

The gap between a demo and production

What a prototype proves versus what a system in daily use actually needs. Usually far less about the model than anyone expects going in.

Get in touch

Either works. Email tends to get a longer reply. I read everything, and reply to most of it.