Problem

Os 3 Bottlenecks that Mandor solves

For boutiques and M&A offices that advise on the sale of a company, acquisition and preparation of assets to go to the market.

1. The partner is the reading bottleneck

For the partner. Those who have the criteria to read contracts, DRE and liabilities are the scarcest person in the house. Mandor makes the first reading of the entire documentation and delivers the opinion by area: the partner stops looking for information and begins to review and decide.

2. Every deal starts over from scratch

For the office and the team. The rule of the previous deal is not written and comes out with whoever leaves. Mandor keeps in the house the memory of the asset, the criteria validated by the partner and the reasons for each refusal.

3. The number does not return to its origin

For the negotiating table and committee. When they ask where the adjusted EBITDA came from, the argument goes off without merit. In Mandor, each number points to the source document, and each section is validated by the team before leaving.

Solution

Mandor is a cognitive network that does the heavy lifting of the deal and leaves the decision up to the office.

Read everything. The advisor uploads the documents and the Mandor reads the entire material, without cutting out excerpts.

Analyze by area. Ten specialized readings (financial, legal, valuation, KYC and others) generate the opinion. The numbers are calculated by formula and each one points to the source.

The team validates. Each section undergoes team review: AI Draft, Under Review, Validated, or Contested.

Save the memory. The criteria validated by the partner and the reasons for each refusal remain at the office for the next deals.

It connects to the market. Invest Match generates the thesis and suggests potential buyers, based on the office base and on the Map with CVM data.

The partner stops reading the document and starts to decide.

Business model

Business model: pay-per-use, with no monthly fee (your decision of 11/08/2026)

The firm buys analysis credits worth for 12 months. The proposed prices are still waiting for your confirmation:

Independent analysis: R$ 890

Packages: 10 thru R$ 7.900, 25 thru R$ 17.900, 50 thru R$ 29.900

Pilot of 30 days: R$ 9.500, converted into credits

Implementation of the “House Criterion”: R$ 19.500, Charged once

Enterprise: tailor-made proposal, with minimum volume per year

Expected average ticket (estimate):

Small boutique: R$ 8 the 10 thousand a year (pilot or package of 10)

Medium boutique: R$ 25 the 50 Not a thousand 1Year 1 (deployment plus package of 25 or 50)

Market

Global market

Closest reference (data rooms): US$ 3,2 the 4,8 billions in 2025, growing from 7% the 22% per year, according to the consultancy.

AI in banks: US$ 15 the 34 billions in 2025. It includes much more than M&A, so it doesn't serve as a direct TAM.

Competitors

Mandor's main competitors today

Internal AI in large boutiques. They have their own AI department and prefer to build at home. It was the reason for G5's refusal.

The analyst with ChatGPT or Claude. It's the most common substitute in medium and small boutiques. It doesn't hold criteria or track numbers.

Global financial AI (Rogo, Hebbia, AlphaSense). Expensive, in English and without a Brazilian standard.

Data rooms and CRMs with AI (Datasite, DealCloud). They only cover part of the process.

Legal AI (Harvey, Jus IA). Just contracts and liabilities.

Competitive differentiation

Mandor's competitive advantage

Made for Brazil. It knows the local standard (CVM, CADE, Revenue) and works by type of transaction: M&A, asset preparation, FIDC, CRI/CRA and structured credit. Global tools don't do that.

Number with origin. Each number goes back to the document it came from. ChatGPT does not guarantee this.

Memory of the house. The discretion of the partner and the reasons for the refusals remain in the office. Generic AI forgets everything with every conversation.

Own market data. The Map crosses the analysis with the CVM's public register: managers, FIDCs, FIPs and transferors. An in-house AI team rarely sets this up.

Human control. Each section is validated by the team, and the decision is left to the office.

Cost. The boutique gains AI capacity without having to hire and maintain its own team.

Entry barrier

What creates a barrier to entry for Mandor

The accumulated memory of the house. Each deal leaves Mandor at the discretion of the partner, the validated corrections and the reasons for the refusals. To change suppliers is to lose that collection. It's the strongest barrier, and it grows with use.

The treated market data. The Map aggregates and corrects the CVM registration: managers, FIDCs, FIPs and related transferors. The data is public, but the treatment took months and is not copied quickly.

The Brazilian standard within the product. The CVM, CADE and IRS rules are in the code and are subject to verification. Those who arrive later must redo this work.

Traction

Nope

Entrepreneurs

Renan Luciano Regonato

Desenvolvedor

Tiago Netto

Comercial

  • HeadquartersBauru
  • Founded08/2026