Ex-CS & Data engineer turned PM. I built Browser and Computer Use agents while also working on Lenovo's global on-device AI project, Qira. I also wrote evals, guardrails, and tuned the agents for accuracy. Then I stepped back for the pricing and the revenue case. Right now that's on-device agentic AI at Lenovo & Motorola Mobility.
I started in code as a CS engineer, analyst, and developer and then moved into product because I wanted the whole picture, not just my slice of it while developing. My experience in engineering means I can get into the evals and guardrails with a team; the PM half means I care whether it ships and whether it pays for itself. Right now both go into Qira, Lenovo's on-device AI.
Five roles across AI, enterprise, and analytics — each with outcomes I can put a number on.
Building agentic AI for Qira, Lenovo's global on-device AI project — three agents (App, Browser, and Computer Use), taken from prompt design through evals and guardrails.
Launched a customization engine that lets customers visualize and preview diamonds before they buy.
Ran product strategy and market research for a new product line, and built the customer app behind it.
Automated Basel III risk reporting with SAS Viya and SQL, replacing a slow, manual process.
Built ML models for lead scoring and fraud detection — better leads for sales, and fewer fraudulent ones slipping through.
Twelve things I’ve built and broken down — AI tools, smart hardware, product teardowns and go-to-market strategy.
An AI learning platform that drops you into real workplace simulations — budget cuts, timeline slips, stakeholders pushing back. No sugar-coated case studies.
View project →
The world's first smart toy for playing Red Light, Green Light with your family.
View project →
Bollywood charades for the living room — act out film titles and get everyone guessing.
View project →
Gamifying financial literacy by applying the $30B fantasy-sports model to the stock market.
View project →
A strategic teardown of Stack AI — where an automation toolkit fits in regulated industries.
View project →
A strategy read on Sensata, an automotive sensor maker, as its market shifts to EVs.
View project →
How Duolingo won at marketing — and where it's leaving upside on the table.
View project →
A strategy and competitive read on DAS Technology, an automotive data and CRM platform.
View project →
A product teardown of Greenlight — what they offer, what the metrics say, and whether it holds up.
View project →
A data-viz dive into how Formula 1 grew into a global sport — and a serious business.
View project →
A go-to-market study of Heron Data, an automation startup — how you actually get from 1 to 100.
View project →
How Ford's business model canvas let it run away with the car market in the early 1900s.
View project →An AI learning platform that drops you into real workplace simulations — budget cuts, timeline slips, stakeholders pushing back. No sugar-coated case studies.
View project →The world's first smart toy for playing Red Light, Green Light with your family.
View project →Bollywood charades for the living room — act out film titles and get everyone guessing.
View project →Gamifying financial literacy by applying the $30B fantasy-sports model to the stock market.
View project →A strategic teardown of Stack AI — where an automation toolkit fits in regulated industries.
View project →A strategy read on Sensata, an automotive sensor maker, as its market shifts to EVs.
View project →How Duolingo won at marketing — and where it's leaving upside on the table.
View project →A strategy and competitive read on DAS Technology, an automotive data and CRM platform.
View project →A product teardown of Greenlight — what they offer, what the metrics say, and whether it holds up.
View project →A data-viz dive into how Formula 1 grew into a global sport — and a serious business.
View project →A go-to-market study of Heron Data, an automation startup — how you actually get from 1 to 100.
View project →How Ford's business model canvas let it run away with the car market in the early 1900s.
View project →The same stack I build, test, and ship with. I live in AI IDEs and model playgrounds, then run delivery on the tools every product org actually uses.
Open to AI product roles and collaborations. The fastest way to reach me is email.