What it is
EvenUp is the category leader in AI for plaintiff personal-injury law — the platform that turns a stack of medical records and case facts into a settlement demand package. It was founded in 2019 in San Francisco by Rami Karabibar (CEO, ex-Waymo), Raymond Mieszaniec (COO), and Saam Mashhad, a practicing attorney who runs product and legal operations. It has raised $385M in total, capped by a $150M Series E led by Bessemer Venture Partners in October 2025 at a valuation above $2B — double its valuation a year earlier. EvenUp says firms run roughly 10,000 cases a week through the platform, that it has helped resolve more than 200,000 cases, and that over 2,000 firms — including 20% of the top 100 U.S. personal-injury firms — use it.
The engine is the Claims Intelligence Platform, built on EvenUp’s proprietary Piai model and trained on hundreds of thousands of injury cases and millions of medical records. That dataset is the moat. The pitch is not “an LLM that drafts demands” but a model that knows what comparable cases settled for.
How plaintiff firms use it
EvenUp anchors on the demand package and expands outward from there:
- Demand letters (Demands). The flagship: a court-ready demand with a narrative of the human impact, medical records organized into exhibits, and comparable verdicts and settlements to anchor the damages number. This is the workflow that justifies the spend.
- AI Drafts Suite. Beyond demands, it drafts complaints, medical-record summaries, negotiation sheets, and responses to interrogatories in minutes rather than the hours a paralegal spends per file.
- Medical chronology and bill summaries. Records condensed into a chronology and a bill summary — the step that otherwise eats a paralegal’s afternoon per case.
- Smart Workflows. Case-lifecycle prompts: send the demand at the right moment, follow up on treatment gaps, flag missing documentation before it costs settlement value.
A defining detail: EvenUp pairs the AI output with human review from its own legal experts before drafts come back to the firm. That raises quality, and it is also why the cost structure looks more like a service than pure software.
Pricing
EvenUp does not publish fixed pricing. In May 2025 it added per-case pricing — one cost per case for the full platform — alongside its seat-based contracts; deals are demo-gated and quoted on firm size, case volume, and onboarding. Public market benchmarks for this class land around $200–500 per user per month, and per-case reports put a demand near $300, climbing to $500–800-plus once add-ons are counted. Treat those as estimates — the real number arrives after a sales conversation. For a high-volume PI firm the math is throughput-driven: hours of workup removed per file against settlement value gained, not the headline rate.
Best for
- High-volume plaintiff personal-injury firms whose bottleneck is demand-package assembly and medical-record review.
- Firms that want the largest PI-specific dataset behind their damages number, with human review baked in.
- Shops standardizing demand quality across associates rather than relying on a few strong drafters.
Alternatives
- Eve — pick it when you want the full intake-to-litigation lifecycle (voice intake, discovery in both directions, a nightly value-driver auditor), not just demand packages. See the Eve entry.
- Supio — pick it when massive medical-record sets are the constraint and chronology extraction depth is the whole game.
- Precedent — the fastest-growing entrant, pitching lower cost and faster demand turnaround; pick it when EvenUp’s price or service-layer overhead is the blocker.
- ProPlaintiff — the budget option, roughly $99–249/user/month, for smaller firms that want demands and chronologies without an enterprise commitment.
Watch-outs
- The output is a draft, not signed work product. A wrong record citation or an inflated comparable in a demand is the firm’s liability, not the vendor’s. Guard: keep attorney review mandatory on every demand, and verify the comparable verdicts against their source before the number goes to an adjuster.
- Cost is opaque and can balloon. A ~$300 base per demand can land at $500–800-plus once add-ons and the human-review layer are counted, and seat pricing is not published. Guard: get the per-case price, what the review layer includes, and the add-on list in writing, then benchmark against Eve and Supio quotes for your firm’s volume.
- You are feeding PHI into a vendor’s model. The corpus is protected health information, and the model improves on aggregate case data. Guard: confirm HIPAA terms, whether your data trains the shared model, and data-export rights before signing — and keep the first term short, given how fast the category and its pricing are moving.