View selected work

Product judgment / Experience design / AI-enabled systems

James Lythe · 2026

Decide what
gets built

Senior product leadership · 001

Product · Systems · Strategy

I help leadership teams turn an ambiguous idea into a product that works. I clarify the problem, define the smallest worthwhile bet, and shape the experience and system needed to deliver it. My current focus is helping AEC leaders decide where AI belongs in their business systems, what to build, what to buy, and what to ignore.

  • Disney
  • NASCAR
  • Britt, Peters and Associates
  • Mattel
  • Hasbro
  • CANY

Selected work

Three products in development, each shown the same way: the problem, the bet, the system behind it, and where AI stops and a person decides.

RFP Extract review workspace showing extracted fields alongside source evidence
01In development · 2026

RFP Extract

A proposal review system shaped around trustworthy evidence, human approval, and clear product boundaries.

  • Workflow architecture
  • Human approval
  • Controlled generation
Read case study
Keystone knowledge workspace with source-aware search and recently updated records
02In development · 2026

Keystone

An internal knowledge interface built on approved records, visible evidence, and honest knowledge gaps.

  • Trust model
  • Source-backed answers
  • Owned knowledge gaps
Read case study
Pursuit Gate decision inbox showing an explainable pursue-as-prime recommendation
03In development · 2026

Pursuit Gate

A decision inbox that turns procurement signals into focused, explainable pursuit choices for AEC firms.

  • Decision model
  • Hard eligibility gates
  • Evidence and gaps
Read case study

How I work

I close the gap between leadership intent and a product that works.

After 15 years in product design, I work beyond the interface. I clarify the customer and business problem, choose the bet, and shape the product and system around it.

I work with leaders and lean engineering teams through delivery. AI speeds exploration, prototyping, and specification. Judgment determines what is worth building and whether it meets a real need.

Screens still matter. So do the parts a user never sees: data, permissions, workflows, automation, and the rules for when a person must decide. Polish cannot fix weak inputs or unclear authority.

01Product judgment + strategy02Experience design + usability03Design systems04AI exploration + prototyping05Specification + delivery guidance06Systems + automation07Build vs buy decisions08Leadership + engineering partnership

Deciding what to build next? Talk to me before you hire developers or buy software.

Contact