Happy Friday!
For the last 20 or 30 years, when discovery was always about moving a client from one well-known system to another well-known system, you had a playbook.
One SI I talked to last week said, “We don’t need to spend hours and hours and hours. It’s fundraising. You already know the requirements.”
In situations like these, repeatability is the name of the game: it’s how you grow margin, make projects faster, and build in best practices for your clients.
But I’m having more conversations with people working on a different kind of project: an AI project. It’s not something they’ve done 100 times - and even if they have, every client and every situation is different. It’s more generalist consulting, where you really have to dig to uncover problem statements, rather than knowing most of it and filling in details.
Discovery matters more than ever in these projects - you need very expert, human consultants who know how to “discover” requirements, not just “gather” them.
If you’re doing discovery on an AI project and you’re feeling like you’re out in the wilderness, I’d love to hear what you’re doing differently. Let’s chat!
What We Shipped This Week & What's Coming Next
This week:
Requirement or decision? - when something lands in the wrong bucket, it’s easier to edit and move decisions, constraints, and risks.
Friendlier project setup - clearer error messages, better validations
Coming next:
We’re working on a slew of UX improvements, thanks to the audit from our friends at Minimal, so expect more delightful workflows and navigation in the next few weeks.
Weekend FOMO
Were you at TSIA this week? I want to hear all about it!
Have a great weekend!
- Ali
