Enter

What it is

Enter is an AI-agent platform for companies that get sued constantly, not occasionally. It reads every incoming lawsuit, pulls in outside evidence, flags claims that look fabricated or duplicated, recommends whether to settle and for how much, and drafts the defense. Attorneys review and validate each answer before it is filed. The company is based in São Paulo and was founded in 2023 by Mateus Costa-Ribeiro (CEO), Michael Mac-Vicar (CTO, formerly CTO of Wildlife Studios) and Henrique Vaz (CPO).

The market explains the product. Brazil carries roughly 75–80 million pending lawsuits, depending on which count you use, and its labor courts handle a share of the world’s labor cases that has no parallel elsewhere. A bank, airline or marketplace operating there can receive thousands of near-identical consumer and labor claims a month. Most legal AI in this catalog is built for the opposite shape of work: few matters, each one large. Enter is built for many matters, each one small, where the economics are decided by cost per lawsuit and win rate across the whole docket.

It is also defense-side. Darrow, Eve, Supio and EvenUp all work for plaintiffs. Enter works for the company on the receiving end.

What you actually get

Two vertical products plus an operating layer:

• Consumer litigation — agents read the case file, enrich it with external records, run fraud checks, and generate a defense tailored to the specific claim rather than a template answer.
• Labor litigation — agents calculate the liability a claim actually carries, build a defense from the claimant’s own data, and prepare hearing arguments at volume.
• EnterOS — the operating layer for the legal department: intake, tagging, tracking of injunctions and rulings, and feeding outcomes back into the next case’s strategy.

The vendor describes more than 30 automated fraud checks, aimed at abusive-litigation patterns such as the same claim filed in parallel courts (lis pendens) and tampered documents. Intake runs through API connections to the courts, legal ERPs and the customer’s internal systems; the agents read contracts, bank statements, audio and video. Security claims on the site: SOC 2, ISO 27001, ISO 27701, LGPD compliance and OpenAI zero-data-retention approval.

Scale, as reported at the Series B in May 2026: more than 300,000 lawsuits a year across 45-plus enterprise customers, including Nubank, Bradesco, Mercado Livre, Airbnb, LATAM Airlines and Azul. Enter says Nubank’s win rate rose 6 percentage points and that LATAM runs 98% of its global litigation through the platform. Those are vendor case studies, not audited figures.

Funding and vendor stability

A $35M Series A at a $350M valuation, led by Founders Fund and Sequoia, in September 2025. Then $100M-plus at a $1.2B valuation on 5 May 2026, led by Founders Fund with Sequoia, Ribbit Capital, Kaszek, Atlantico and ONEVC — tripling the valuation in eight months. Revenue reportedly grew more than 10x after the Series A and the customer base tripled; headcount was reported at about 100, scaling toward 150. A no-acquisition, no-shutdown check on 1 October 2026 turned up nothing. Vendor risk here is concentration on one jurisdiction, not survival.

Pricing reality

No public price, no tiers, no self-serve. Reporting at the Series B describes a hybrid model: roughly 30% of revenue from success-based fees, the rest billed upfront for the technology. That is the right structure to negotiate around, because a success fee means the definition of “success” is the contract.

Pin three things before signing: how a win, a favorable settlement and an avoided claim are each defined and priced; the per-lawsuit or volume band the upfront fee covers and the overage rate past it; and who pays the partner or in-house attorneys who validate each filing — that review cost sits outside the platform fee and will not show up in the demo.

Best for

General counsel and legal-ops leads at consumer-facing enterprises in Brazil — banks, fintechs, airlines, insurers, marketplaces, retailers — carrying thousands of repetitive consumer or labor claims a year, where the current model is a panel of mass-litigation outside-counsel firms billing per case. The pitch is the cost per lawsuit and the win rate, measured across the docket.

Not for you if your litigation is a few dozen bespoke matters a year; the platform earns its fee on repetition. Not for you if your exposure sits mainly outside Brazil: an international version is mentioned on the site, but every published customer result is Brazilian. And not for you if you want a drafting assistant for your own lawyers rather than a system that runs the docket.

Alternatives — and when to pick them instead

• Mass-litigation outside counsel (the “contencioso de massa” firm panel) — the status quo Enter displaces and the largest share of spend in the segment. Keep it when your volume is moderate, your win rate is already acceptable, and you would rather pay per case than redesign intake.
• Thomson Reuters Legal One and Projuris — the incumbent Brazilian legal-ERP layer by installed base, used for case tracking, deadlines and spend. Pick one when the gap is control and reporting over a docket your firms already defend; they track litigation rather than draft the defense. Expect Enter to sit on top of one of them, not replace it.
• Legora or Harvey — the fastest-growing general legal AI platforms, valued at roughly $5.5B and $11B respectively in the Series B coverage. Pick one when your work is complex, high-value matters where a lawyer drives each document; neither is built to run hundreds of thousands of near-identical defenses end to end.
• Jurídico AI — a Brazilian entrant that launched a labor-litigation management and intelligence platform for legal departments. Shortlist it if labor claims are the whole problem and you want a second bid in the RFP.

Watch-outs

• Every performance number is vendor-supplied. The win-rate lift, the 98% LATAM figure and the cost reductions come from Enter case studies. Guard: run the pilot on a closed back-catalog — feed it a few hundred lawsuits you already resolved, have it recommend settle-or-defend blind, and compare its calls and settlement values with the actual outcomes before any live filing.
• A success-fee contract makes the definition of a win the most important clause. If a settlement below the claimed amount counts as a win, the fee can trigger on cases your firms would have settled anyway. Guard: baseline your current win rate and average settlement per claim type before go-live, and price the success component only against improvement over that baseline.
• Fraud flags are accusations. Labelling a claimant’s filing as fabricated or duplicated in court carries professional and reputational consequences if the flag is wrong. Guard: require that every fraud-based argument carries the underlying evidence in the file and a named attorney’s sign-off, and track the rate at which courts accept those arguments as a pilot metric.
• Single-jurisdiction depth. The models, the fraud checks and the customer proof are tuned to Brazilian consumer and labor procedure. Guard: if you operate across Latin America or the U.S., scope the contract to Brazil and require a separate, evidenced pilot before extending it to any other court system.
• No public API docs or MCP server. Integration happens through Enter’s own connectors to courts and legal ERPs. Guard: name your legal ERP and your outside-counsel firms in the demo and make the vendor show the intake and filing handoff live, including how a case moves back to a human-run firm when you pull it from the platform.