BUILDER + TRANSLATOR
I turn ambiguous business intent into shipped AI-native products. I design the agentic systems, write the specs, run the evals, and build the thing — across product, AI, and mobile. I speak both Engineer and Executive, which is how the two stay in sync.
LOADOUT
Equipped_SkillsAI & DELIVERY SYSTEMS
PRODUCT & DELIVERY
MOBILE & ENGINEERING
TOOLS
EXPERIENCE_LOGS
Loading_History...QUANTITATIVE FOUNDATION
Built complex trading models for Deutsche Bank. Developed a rigor for numbers, financial structuring, and large-scale data analysis.
GLOBAL DECISION TOOLS
Took four interactive decision tools from 0→1 for World Bank teams and government ministries worldwide — ran the user interviews, wrote the specs, scoped each one, then built and deployed them myself in d3.js. Product work, end to end.
0→1 PRODUCT DELIVERY
Led a team of developers and designers to ship resourceprojects.org — owned it end to end, turning ambiguous stakeholder needs into structured requirements and shipping the alpha in three months. Built an automated pipeline that cut processing cost 20x and turned 16,000+ rows of non-machine-readable PDFs into structured data.
PRODUCT × ENGINEERING
Partnered with product and engineering on positioning and launches for a real-time analytic database. Built the developer-education program from scratch — 10 courses, 50+ videos, 25 GitHub example repos — and delivered a full rebrand and funnel at 1/10th the prior budget.
AI-NATIVE BUILDER_MODE
Shipped an AI-native test-prep platform to iOS and Android, leading a distributed team of engineers and designers. Built an AI that turns official exam syllabi into high-quality study material and graded assessments — source-grounded, rubric-scored, with human-in-the-loop QC across 36 curriculums incl. FINRA SIE/Series 7/63. Redesigned the app hands-on with Claude Code.
AI SYSTEMS
Built an AI GTM-intelligence system, 0→1: it discovers ~12,000 US water utilities, reads their public documents with a cost-aware model cascade, and surfaces which are about to buy — behind a Next.js product with a grounded RAG assistant, evaluated against a human-curated gold set.