Every large company spends millions a year on software, usually the second biggest cost line after people, and negotiates it against vendors who close thousands of deals a year with list prices that mean nothing. Vera AI is the platform that puts the market data, the preparation, and the negotiation machinery on the buyer's side of that table.
This page explains the thesis, what is already built, and why we believe the moat compounds. If it resonates, request the investor brief at the bottom.
Enterprise software pricing is deliberately opaque. Discounts off list run from 10% to past 70% for the same product, the real numbers live inside vendors' deal desks, and the buyer's only reference point is usually their own last contract. The result is a persistent, structural overpayment that finance teams can feel but cannot prove.
The traditional answers do not scale. Consultants are excellent and expensive, and they arrive one deal at a time. Spend management tools show what you spend, not what you should pay. The gap in the market is the reference data plus the analyst work, delivered as software, on every deal.
Years of compounding uplifts, license model changes, audit campaigns, and now per seat AI pricing pushed software from an IT line item to a CFO agenda item. Budget owners are actively looking for leverage for the first time.
The preparation that used to take a consultant a week, decoding the contract, placing the deal against the market, writing the playbook, now runs in minutes on frontier models, grounded in our reference data and checked before it ships. The economics of buyer side advice changed by an order of magnitude.
Procurement is an early, obvious home for AI agents negotiating with AI agents. We publish an open Agent Negotiation Protocol and already run agent led negotiation workflows, so the platform is positioned for that shift rather than disrupted by it.
Vera AI is live at app.vendorbenchmark.com: benchmarking against real closed transactions, AI negotiation playbooks and dossiers, a research library, contract management with semantic search, Vera the portfolio analyst, self serve optimization tools for the major vendors, and autonomous workflows that watch renewals, invoices, and market drift around the clock. The features catalog lists more than fifty shipped capabilities, and the in-app release notes show the shipping cadence: substantial features land weekly, not quarterly.
Benchmark cohorts built on more than 500,000 real closed transactions, curated and graded by analysts, plus a maintained price book with weekly automated verification against vendors' own published pages and a quarterly Software Price Index. This corpus took years of buyer side engagements to assemble and is not reproducible from surveys.
An anonymized outcome network on a give to get model, buying blocs that pool negotiation signal, a peer network, and vendor scorecards. Every negotiated outcome that flows back makes the next customer's benchmark sharper, which is the classic data flywheel, applied to a market that never had one.
The platform runs the whole job, from the first price check through the negotiation war room to invoice verified savings proof. Depth like this is slow to copy and sticky to leave: the contracts, the renewal calendar, the audit trail, and the team's working history all live here.
We take no vendor money and sell nothing to the sell side. In a market where most data providers also serve vendors, a clean buyer side position is a durable brand asset that customers actively select for.
AI does the volume, humans sign the verdicts. Every AI figure is grounded in stored data and citation tagged, and delivered benchmarks carry analyst review. That quality bar is what lets a CFO forward our reports to the board, and it is hard for pure software competitors to match.
An open Agent Negotiation Protocol, a public API and MCP surface, and agents already negotiating inside the product. When procurement becomes agent to agent, the side with the market data wins, and we intend to be that side's infrastructure.
The core is straightforward B2B SaaS: a 30 day trial, a Standard plan at $5,000 per month, and Enterprise agreements with analyst benchmarks, SSO, and advisory sessions included. On top of the subscription sit revenue lines that scale with the value delivered rather than the seat count.
Recurring revenue with strong expansion mechanics: more contracts under management, more seats, more AI usage, and plan upgrades as the platform becomes the team's system of record for vendor spend.
Our team runs the renewal end to end, and fees are charged only on savings verified against the invoices that follow. Aligned incentives, high margin analyst leverage through the platform, and a natural upsell from self serve.
For qualified deals we underwrite a savings floor with our fee at risk, settled on invoice evidence. Underwriting is gated by the same benchmark data that powers the product, which is what makes the guarantee safe to offer.
Expert advisory sessions beyond the plan quota, a partner program for advisories and resellers who bring their clients onto the platform, and benchmark credits for one off needs round out the model.
Vera AI is built by Redress Compliance, the independent buyer side licensing and negotiation advisory enterprises bring in when the Oracle audit letter lands or the Microsoft renewal doubles. The platform is that practice turned into software: the data from real engagements, the playbooks from real negotiations, and the analysts still in the loop on every delivered benchmark.
That origin matters for investors. The product was not imagined in a pitch deck; it automates work customers already pay for, in a category the founders have practiced for years.
Request the investor brief: the market, the model, the metrics, and the roadmap.
This page is provided for general information only. It is not an offer to sell or a solicitation of an offer to buy any security, and nothing here should be read as investment advice. Figures describe the product and its data as shipped and are not projections of financial performance.