Enterprises put a lot of work into negotiating contracts: pricing, volume tiers, service guarantees. Once the deal is signed, the contract usually gets filed away, and much of that negotiated value slowly leaks back out.
A new feature on MSN, The $3 trillion problem hiding in corporate contracts, and the AI built to find it, tells the story of how Terzo was built to find that money.
Money that leaks, not money that's lost
Across the global B2B economy, trillions of dollars are exposed to leakage between what companies negotiate and what they actually pay, use and recover. None of it is stolen, and none of it disappears in a market crash. It leaks out through:
Invoices that creep above contracted rates
Renewals that fire on autopilot
Rebates nobody claims
Warranties nobody enforces
A 95-slide wake-up call
Brandon saw the problem up close at Microsoft, where he managed nine-figure enterprise deals. In 2018, a global automaker asked him to summarize its Microsoft agreements. The result was a 95-slide PowerPoint, with one contract screenshot for each country.
"I'm in meetings over here talking about sending Teslas to Mars," he told the publication. "And then I'm over here building a PowerPoint with ninety-five slides for contract data. This just doesn't make sense."
He had spent a decade at Oracle, IBM and Microsoft watching the same pattern repeat. In 2020 he founded Terzo on a simple idea: contracts aren't legal paperwork. They're financial assets.
From contract intelligence to financial intelligence
Terzo turns contracts, amendments, purchase orders, invoices and supplier data into a structured financial intelligence layer. It then continuously compares what an enterprise agreed to with what it actually buys, uses and pays for.
For example, if a master services agreement caps annual price increases at 3% and an invoice arrives 8.2% higher than last year, Terzo flags it, attaches the evidence, calculates the dollar impact and routes it for recovery.
As Brandon puts it, this is a scale problem, not a people problem. Auditors can only sample a small share of transactions. Software can review all of them.
Why the data is the moat
Large language models have made it easy to summarize a contract, but summarizing isn't the hard part. The hard part is creating trustworthy, structured financial data from millions of messy documents. The article cites a Gartner prediction that through 2026, organizations will abandon 60% of AI projects that aren't supported by AI-ready data.
Terzo's approach keeps humans in the loop. AI extracts every clause, obligation and SKU, and human reviewers validate the output to a contractual accuracy guarantee above 99%. To date, Terzo has processed more than two million contracts representing over $500 billion in enterprise agreements.
Results at Fortune 500 scale
$300 million saved across procurement at one Fortune 500 company, where a senior procurement leader calls Terzo "the system of truth for tracking $8B of spend in my org"
$25 million saved in IT spend in another company's first year
$250 million in verified savings across a third company's supply chain
Terzo doesn't replace the ERP or the contract repository; it connects them. Deployment takes about a month, the client commitment is roughly an hour a week, and first findings typically arrive within 90 days.
Finding the money you can't see
As enterprises rush to deploy AI agents, Terzo is betting that the winners won't be the companies with the most chatbots. They'll be the ones with the cleanest financial data and the clearest link between AI and measurable results.
"Every enterprise has money it can't see," Brandon said. "Our job is to find it."
Read the full feature on MSN →
Want to see what's leaking out of your contracts? Book a demo
