Enterprise APIs, from roadmap decisions to customer adoption.

Prioritizing what customers needed to go live

Customers could integrate against the Test API while live markets were still being validated. I kept its request and response contract aligned with production, with live activation and customer adoption as separate stages. One operator needed alternate lines, or different thresholds for the same statistic, before launch; another was already live.

Following an integration issue through customer-confirmed recovery

I lead client-facing resolution of NBA Fast Props API integration issues. Fast Props provide short-lived betting opportunities during a game, so disappearing selections directly affect the customer’s live product.

When one customer reported selections disappearing, I inspected their requests in Datadog and investigated market grouping and newly added player state. I coordinated staged fixes with Engineering and stayed with the issue until the customer confirmed that selections had stopped disappearing.

My scope

I set roadmap priorities, define product requirements and API behavior, and lead cross-functional delivery and customer follow-through. Engineering and Data Science build the production services and models.

For NBA alone, I owned 10+ product releases between November 2024 and January 2026, spanning new market families, shared pricing, results and client controls.

Ship three, hold one

Four existing NBA markets were proposed for live combined-bet pricing. Team-points calculations did not reconcile with game totals. I shipped three ready markets and held back team points, accepting a smaller release so customers could use the ready scope without exposing the inconsistent fourth.

The April 15, 2025 release made first-quarter player points, assists and rebounds available in combined requests during a game. These were existing markets gaining a new way to be used, not three new models.

Give a new product the results it needs

NBA Leaders ranks players within a team or game. I defined API and results behavior for its October 2025 release across 16 markets. When combination-rank fields were missing upstream, I confirmed that they needed to be supplied and coordinated Data Engineering’s additions. Adding separately ranked values would not produce the required combined-stat ranking.

I also specified distinct treatment for a player who did not participate. The shipped capability connected pricing, combined requests and result fields so customers could settle these markets using the corresponding rankings.

Resolve customer requirements in the shared API

In a separate NBA release, I removed shared combination restrictions that blocked a customer’s integration. For CFB, I expanded the shared pricing and settlement APIs to support 26 existing team/match markets in January 2026. Customers could retrieve results through established APIs. We reused request and response structures; customers still needed to configure and test the additions.

More granular availability, without breaking the existing contract

A market-level suspension pauses every selection together. I led optional selection-level controls so clients could keep an eligible side open, such as the over, while the under was suspended. Clients could opt in when ready, while existing integrations kept their market-level handling.

I introduced selection-level suspension fields across pre-game and live pricing APIs in November 2025. In January 2026, I led their activation for live props, live alternates and in-play composition.

{  "market": "player_points",  "line": 24.5,  "suspended": true,  "selections": [    {      "side": "over",      "suspended": false    },    {      "side": "under",      "suspended": true    }  ]}