Engineering the Future of MLB Pricing: A Projection Pipeline Overhaul

At Huddle, we're constantly pushing the limits of pricing technology to help sportsbooks deliver sharp, scalable, and efficient betting markets. With the MLB season demanding pricing accuracy across thousands of games and hundreds of players, we knew it was time to reimagine our projection pipeline from the ground up.

Why the Overhaul?
The old system did the job — but it came with high maintenance costs, redundant workflows, and limited flexibility. It relied on outdated spreadsheets, legacy infrastructure, and manual trader inputs that no longer aligned with the speed and scale required for MLB.

What Changed?
Our engineering team, led by Ivan Juraga, took on a complete architectural revamp. Key improvements included:

• Migrating from XLSM files to Snowflake-based CSV storage.
• Replacing Windows VMs with leaner, cloud-native processes.
• Enhancing data validation, replay capabilities, and integration pathways.
• Integrating our proven passing param framework to significantly reduce manual overhead.

What It Means for Our Partners
The result? Faster deployment, cleaner data, more flexibility for traders, and a pricing engine that's easier to maintain and scale. We're already seeing positive feedback internally — fewer bugs, smoother workflows, and a more responsive system overall.

This overhaul isn’t just a one-off. It’s a blueprint for how we plan to evolve projection models across all sports moving forward.

📄 Dive into the full case study and technical breakdown in our latest whitepaper: