Product leader · Builder · Cleveland, Ohio
I’ve spent more than ten years in product management, connecting complex problems with solutions worth building. These days, I’m building with AI, testing how agents behave, and figuring out what actually helps the person using them.
Selected work / 01–04
A personal project and selected client work.
The decisions, tradeoffs, and open questions.
01 / Personal project · Still learning
I built Rare Tomato so you can review activity from different AI assistants, turn what they got wrong into rules shared across them, and track how each follows those rules over time.
Built and deployed
Read the case study: Teach your AI agents to work your way.
One project: government-affairs strategy
System design + specialist data02 / Government-affairs strategy · System design
I designed a strategy system informed by experienced government-affairs practitioners, the firm’s growing institutional memory, and changing government timelines. The knowledge it needed had no suitable public dataset, so I also built the collection tool in case 03.
Designed; implemented by engineering
Read the case study: AI that uses a firm’s specialist knowledge to build strategy.What I designed
Collect judgment from experienced practitioners where no suitable public dataset exists.
Use the firm’s accumulated knowledge to make each strategy relevant.
Align strategy with shifting state, legislative, and budget calendars.
Check dependencies and keep proposed updates under the user’s control.
Collecting the specialist knowledge needed for the strategy above
03 / Government-affairs strategy · Data collection
No suitable public dataset captured how experienced government-affairs practitioners make strategic decisions. I built a tool to collect their judgment as structured examples for model evaluation and later training.
Working instrument delivered
Read the case study: Turn specialists’ judgment into data an AI model can use.
04 / Client work · Anonymized
I mapped a firm’s compliance process, then designed AI to identify and prioritize filings, assign accountability, and prepare the right forms from institutional memory. People make the final submission.
Designed; implemented by engineering
Read the case study: AI that prepares a firm’s required filings for submission.What I designed
Translate an existing compliance operation into an AI workflow.
Identify required filings and make a staff member accountable.
Use institutional memory to prefill supported information.
Carry the work to the submission boundary, where a person takes over.
AI policy
Graduate coursework · Purdue MS in AI Management and Policy
Frontier AI / Incident reporting
I analyzed California’s SB 53, mapped the stakeholders shaping it, and proposed a Near-Miss Learning Program from the Cal OES perspective.
Statutory analysis · Stakeholder mapping · Policy recommendation
Read the policy case studyCoursework; not affiliated with or endorsed by Cal OES.
About
I’m Lia Martinez, a product leader with more than ten years of experience. My work spans an AI-native startup, a major US bank, and a digital benefits platform where I was the first product hire.
I’ve owned product strategy, designed detailed workflows, written engineering specifications, and led cross-functional delivery. More recently, I’ve been building with AI tools and using evaluations to make product assumptions testable.
Get in touch