• 12 pp increase in win rate on lawsuits managed by Enter, with audit and review by licensed attorneys, 8,000 of which have already closed
• 25% reduction in average damages paid per lawsuit driven by a 13% reduction in average payout per case and a 30% increase in win rate
• +40 dismissals for abusive litigation, 2 sanctions for bad faith litigation and a referral for criminal investigation in the case of an abusive litigant with many lawsuits, as mapped by Enter's AI agentsWidespread document tampering and fraud detected through the reuse of the same proof of residence across dozens of lawsuits, including one case that resulted in dismissal and a fine of 9% of the claim value imposed on the plaintiff

Payroll-deductible credit is one of the top three most litigated issues in banking lawsuits in Brazil¹. In 2023, the sector saw 585,000 lawsuits on the subject — a 143% increase over three years². By 2024, that volume exceeded 1,700 new lawsuits per day³.

Mercantil manages a large number of lawsuits per year. As bank fraud and abusive litigation grew, the bank needed sharper tools to prevent fraud, detect it early and identify abusive litigants in a timely manner. Specifically, the bank was looking for tools to handle:

• Granular lawsuit classification and root-cause identification of losses — knowing not just that fraud occurred, but also what specific type of scheme was involved and how to improve preventive measures to protect customers.
• Access to an intelligence network capable of flagging abusive litigants with a known history within the system.

The central question was:

In high-volume consumer litigation like Mercantil's, classifying lawsuits introduces operational complexity that manual review cannot handle. Enter connects via API to the Brazilian National Justice Council (CNJ) and Mercantil's legal ERP, reading each lawsuit in detail — identifying the product involved, the allegation and the claim value.

That registry enabled pattern analysis at scale. In a sample of the bank's lawsuits, Enter found that 65% of fraud incidents fall into three main categories.

The finding spurred two parallel tracks: on the legal side, Enter's AI began refining its responses to lawsuits by applying the most relevant legal inputs for each fraud type. On the product side, the legal team fed the anti-fraud unit data that informed investment in more preventive measures — addressing root causes and improving customer security.

O dado alimentou dois caminhos: no jurídico, a IA da Enter passou a refinar as teses utilizando os subsídios disponíveis de forma mais estratégica para cada tipo de golpe.

No produto, o jurídico alimentou o time antifraude com dados que induziram o investimento em mais medidas preventivas, atuando diretamente na causa raiz do problema e trazendo cada vez mais segurança para seus clientes.

Enter's AI agent understands Mercantil's reality and knows which documents are required for each lawsuit type. It reads full lawsuit files, flags any gaps to the partner law firm before the lawsuit answer is drafted and verifies, for example:

• Whether the correct contract is attached
• Whether the biometric signature matches the identity document
• Whether supporting documents are valid and current
• Whether the dossier, even if it spans hundreds of pages, demonstrates that the funds were subsequently utilized
‍

Enter's AI agent runs more than 30 violation checks and builds an attorney dossier,  consolidating the history of lawsuits against the bank, behavioral patterns and connections to other litigants.

In one case, a single litigant filed hundreds of lawsuits against Mercantil in less than 10 months. Enter's AI agents mapped the full set and identified the pattern:

• 97% of lawsuits with identical hearing waiver requests
• 99% with fee waiver requests and no justification
• Standardized claim value of US$1,870
• 13% of the entire litigated base concentrated in 39 CPFs (Brazilian individual taxpayer ID numbers) with four or more claims on the same contracts
• Identification of the referral firm linked to the litigant — located through document analysis, including powers of attorney from the digital signature provider

With the dossier in hand, Mercantil began challenging lawsuits using a targeted legal strategy. Together, the teams secured more than 40 dismissals for abusive litigation from 16 different judges, along with 2 sanctions for bad faith litigation and referrals to NUMOPEDE, a Brazilian judicial agency overseeing legal misconduct.

In the quarter after Enter mapped the pattern, new filings from that abusive litigant dropped by 98% — an estimated US$3,100,000 in claim value that did not enter the pipeline in that single quarter.

The impact extends beyond the bank. When a structured dossier reaches a judge — with documented patterns, connections between litigants and evidence of coordinated conduct — the judiciary gains information it would otherwise never receive in organized form.

Detection at scale does not just protect the bank. It feeds the legal system with organized evidence that allows judges to act beyond the scope of individual lawsuits.

‍

With Enter, we've been able to optimize the use of legal inputs with a strategic approach to consumer litigation, correlating the legal theory, the inputs and the specific type of fraud scheme or allegation. It lets our team operate strategically, increasing wins and reducing litigation costs.

— Luiz Bolognani Head of Civil Litigation and Judicial Collections, Banco Mercantil

With patterns mapped, Enter develops specific legal theories for each lawsuit type, which are reviewed and audited by partner law firms before filing. These include abusive litigation, abusive claim splitting and challenges to allegations of unauthorized loans.

For that last category, the agent reads all legal inputs in full, helps identify patterns in fund use in lawsuits of disputed contracts or fraud, pinpoints the month of signing and traces what happened to the money. In one case, the lawsuit was filed 20 months after signing, and bank statements showed the funds had been transferred to accounts directly linked to the account holder, month by month from the start.

That level of analysis is what allows Enter to build a customized legal theory for each lawsuit.

Enter conducted a ruling analysis to map the main grounds for adverse judgments against Mercantil. The study found that the leading arguments can be countered with legal inputs such as biometrics, geolocation, device profile monitoring and IP address tracking — all of which can be better leveraged through Enter's AI.

Mercantil moved fast with the diagnosis: it addressed the top win-rate detractor by improving how it used the dossier it already had but had not previously applied exhaustively.