Sports.

AI.

LLMs.

Sports.

AI.

LLMs.

Grounded sports-data assistants, agents, and the monetization surfaces around them — built to move a number, not to demo well.

Built recently, independently

What I’ve been building on my own

The work above shipped inside large organizations. This shipped without one.

The monetization stack

A map of every place a sports publisher makes money, the infrastructure underneath it, and twenty-two recurring failures with the mechanism behind each one.

Read the reference →

LLM

A grounded sports assistant that answers from real data, cites its sources, and refuses when the answer isn’t there.

Try it live →

The working library

Reusable skills and connectors for sports monetization — each one specified with its inputs, outputs and the decision it makes.

Open the library →

Where the work

has shipped

Built inside the biggest names in sports media and entertainment — shipping fast, moving revenue, and holding up in production.

PFF · Walt Disney · Fox Sports · USA Today

49%

YoY revenue growth driven at Fox Sports through native video advertising.

2.3B+

$2.3B+ in annual revenue across the sports & entertainment platforms behind this work — Disney, Fox Sports, and USA Today.

Years shipping in sports tech - payments, ticketing, streaming, and data.

Designing, building, and shipping products that move revenue.

What gets built

LLM and AI products that move revenue

LLM-powered products for sports and entertainment, grounded in real data and built around how audiences actually convert.

Sports-data LLMs

Grounded assistants like LLM that answer from real data, cite their sources, and refuse to hallucinate.

Responsible-gambling compliance engines

Betting and DFS monetization with age, state-eligibility, and brand-safety guardrails built in, not bolted on.

ML-driven engagement & retention

Recommendation and lifecycle work that keeps audiences active across the full season, not just on game days.

Dynamic pricing & yield optimization

Real-time pricing and ad-yield systems that lift conversion and revenue per visit.

See it working

LLM runs live in your browser. Ask a sports data question and watch it answer from the data, cite its sources, and refuse when it doesn't know.

Try LLM →