AM

Atiq Israk

Topic

AI Product Management

Most AI products fail not because the model is weak, but because nobody named the business number it had to move. I scope AI where it beats manual reliably — brand-voice wedges, inventory truth, conversational automation — and kill everything else. My AI work spans consumer apps (Anne), SMB platforms (Kaizen), enterprise inventory (AssetIQ), and customer automation (Navbot). The through-line: deploy AI only where operators trust the output enough to change behavior.

Impact

Key stats

264%
Revenue growth (AssetIQ)
85%
Inquiry automation (Navbot)
Core retention (Kaizen wedge)

Frameworks

How I think about this

FAQ

Common questions

How do you decide when AI is worth building?
When it beats manual reliably on a metric leadership tracks — accuracy, response time, retention, or revenue — and operators trust it enough to change behavior.
What's your approach to AI product scope?
Find a narrow wedge with a measurable outcome, ship the smallest proof that earns trust, then expand. Kaizen's brand-voice focus doubled retention vs. a broader launch.
Consumer vs. enterprise AI — what's different?
Consumer AI wins on retention and acquisition efficiency (Anne). Enterprise AI wins on accuracy and auditability first (AssetIQ held the line on reconciliation before dashboards).
Can you share AI product case studies?
Yes — browse Kaizen, Anne, AssetIQ, Navbot, and Pico on the work page, each with headline metrics and narrative.