ContractIQ

The problem

Hospitals sign hundreds of payer contracts, each with its own rate tables, carve-outs, escalators and fine print. Knowing what a given claim should have paid means reading and modeling all of it. When that work is manual, underpayments slip through and rate negotiations rest on guesswork.

My role

I’ve shaped ContractIQ from its earliest design through to production, and I lead the ContractIQ and Pricing & Scoring teams. Pricing & Scoring applies ML to denied and underpaid claims to surface the ones most worth pursuing on appeal. Across both, I set technical direction and build the team.

What it helps us and our clients do

• Find money owed. Compare what was paid against what the contract says should have been paid.
• Answer rate questions in minutes, not days. Model “what if this rate changes?” across a portfolio of contracts.
• Scale expertise. Contract knowledge that lived in a few analysts’ heads becomes something the whole organization and every client can use.

What I learned

Our first users were a small team of contract experts, most of them not technical. The hard part wasn’t the model. It was representing dense contract language in a way people could read, trust and change themselves. We started simple, aiming for adoption with a lightweight prototype. We quickly learned the problem needed something much bigger and bolder, and rebuilt around that.

Explore the idea

The ContractIQ Impact Explorer is an illustrative model built on synthetic data and published industry benchmarks. It is not the ContractIQ interface.