"Hari is an exceptional Product Manager with deep knowledge in the AI space. His ability to bring together competing views and needs into a cohesive product vision is extremely valuable."
EVP, Chief Supply Chain Officer · Healthcare
Case study · AI execution
AI decisioning and execution that survived deployment
30-50% manual work reduction through OCR, route optimization, and workflow redesign — AI gains from deployed workflows, not prototype theatre.
Company details are generalized. I am happy to walk through the specifics on a call.
The situation
Most industrial AI initiatives produce impressive demos and quiet failures. The models work; the workflows reject them. The hard part is rarely model novelty — it is making the work usable under real operating conditions, where data is imperfect and the people affected have veto power in practice.
The mandate here spanned decisioning and execution workflows: document-heavy back-office processes, field routing, and a portfolio of systems that needed AI-ready foundations rather than bolted-on features.
The approach
The AI was aimed at work operators and back-office teams already wanted to shed — OCR for document processing, route optimization, and workflow redesign that removed steps rather than adding a dashboard. Each gain was measured against the deployed workflow, not the prototype.
Advisory work applied the same discipline at roadmap level: pressure-testing where AI belonged in a product portfolio before the bets hardened.
The results
- 30–50% less manual work through OCR, route optimization, and workflow redesign.
- $13M+ ARR portfolio with AI-ready foundations across 10+ systems.
- $100M+ revenue opportunities shaped through advisory work.
What it says about the work
The gains came from deployed workflows, not prototype theatre. AI that removes work people already resent needs no adoption campaign.
More selected proof
Facing a similar decision?
A short note on the AI bet you are weighing is enough to start. I usually reply within one business day.