Map the workflow, constraints, users, data, and adoption risks.
I turn messy work into working solutions.
I combine product judgment, analytics, and hands-on building to take AI workflows from discovery to deployment. Former Apple and Agoda analyst. UCLA Anderson MBA. Co-founder at Apocrypha.
Prototype quickly, connect the data, and design for real decisions.
Deploy, learn from use, and iterate on the operating model.
Where technology meets operations.
Two projects that show how I work: start with a concrete user problem, build the smallest useful system, and improve it against real feedback.
Operational decision support for fuel logistics
Built and sold a centralized ETA workflow for a Thai logistics operator managing more than 300 trucks. The product replaced fragmented GPS and Excel checks with one operational view for 120 to 140 daily vehicle movements.
A resume-aware job search and fit system
Created an end-to-end product that translates a candidate profile into a search strategy, retrieves roles, scores fit, checks sponsorship signals, removes duplicates, and turns results into a reviewable job pipeline.
Open live demo ↗From analysis to ownership.
I began in engineering, moved into analytics, and kept moving closer to the product and the people using it. Each role added a layer: operations, forecasting, product quality, and now applied AI delivery.
Apocrypha AI Services
Lead discovery, product design, and implementation for AI-enabled operational workflows. Built and delivered the company’s first paid product.
Apple
Evaluated recommendation, search, and content systems. Informed a Similar Artists model update that improved accuracy 5% across approximately 93 million users and helped improve Thai search accuracy 30%.
Agoda
Built regression-based demand and booking-maturity models that reduced forecast error 40% and improved staffing decisions across language markets.
Predictive · ABB
Automated digital analytics QA to cut audit review time 30%, then built operational tracking tools that increased service rates 10%.
Build. Decide. Measure. Operate.
Applied AI delivery
Design and ship reliable AI workflows with evaluation and human judgment built in.
Product and discovery
Turn ambiguous needs into a focused scope, usable workflow, and clear iteration path.
Analytics and experimentation
Use data to define the problem, test assumptions, and make the next decision concrete.
Operations systems
Connect tools, data, and people into systems that survive contact with daily work.
UCLA Anderson School of Management
MBA · Entrepreneurship specialization · 2026

Chulalongkorn University
BEng · Mechanical Engineering