Receivables
PT

Catalisa · Platform for financial products

Build your financial product in days — no build-out required.

Origination, decisioning, disbursement and collection on building blocks that already run in production. You design the pipeline; the foundation is already standing. Credit is the first, and the same base serves the next one.

For people who build credit operations: FIDC managers (Brazilian receivables funds) · securitizers · credit fintechs · mid-size banks · credit unions

1 · BackingIt exists, and belongs to whoever claims it
2 · DossierEverything the decision needs to see
3 · DecisionThe rule is yours, and your team changes it
4 · DisbursementGoes out with sign-off and an audit trail
5 · SettlementOnce paid, the receivable settles on its own

The thesis

Eighty percent of the process is the same in any credit operation.

Switch the operation and notice: the five stages don’t move. What changes is the data that decides — and that is your competitive edge.

The reference operation

The whole backbone, proven end to end.

Invoice over WhatsApp Dossier disallowance by insurer Desk above the limit, a person signs Pix disbursement authorized Settlement the insurer pays and the receivable settles
Credit desk · review queue

Awaiting decision

Clínica VegaR$ 320,000above the limit
Instituto AuroraR$ 84,500
Núcleo DiagnósticoR$ 41,200

Clínica Vega3 OS invoices

PayerHealth insurer
Face valueR$ 320,000.00
Automatic approval limitR$ 300,000.00 · exceeded
Disallowance coverage63%

The policy cutoff is 80%. The third invoice arrived without the payer’s CNPJ (Brazilian company tax ID), and without it there is no history for that insurer.

Approve with conditionsDecline

Illustrative names. The screen is the real one — and the person pressing the button is a human.

R$ 320k

the sum of the three invoices in the demo

R$ 300k

the approval limit — which is why the deal goes to the desk

63%

the disallowance coverage after the third invoice arrives without a CNPJ

The script was designed to teach, not to impress: a demo that always approves hides exactly the step where a person decides.

Next step

Thirty minutes with one of your receivables.

You bring a sample portfolio. We load your policy. The proposal comes out on screen right in front of you — the same one your customer would see.

The missing layer

A tailor-made product, in days.

The market sells you the top floor ready-made — and someone else’s process comes with it. Or it sells you the basement parts, and assembling them becomes an IT project. What’s missing is the middle.

your policy — you move it, no code Risk cutoff how far it approves Approval limit when it goes to the desk Pricing how the cost is built The proposal what the customer reads the foundation — already running in production backing · dossier · decision · disbursement · settlement OS invoice advance the reference operation Supply-chain finance other settings, same pipeline Card receivables other settings, same pipeline

The ready-made top floor

Quick to switch on. But the process is the vendor’s, and your competitive edge becomes theirs.

The basement parts

Great parts. But assembling the process is still an IT project, and the second operation costs as much as the first.

The middle, which is what’s missing

A ready foundation, a product of your own, and the middle layer in the hands of people who understand credit — not of people who write code.

The third way out, and what it costs

Some solve this by outsourcing the whole pipeline.

It’s the model that has dominated large-scale credit in Brazil for more than a decade: the platform runs under the bank’s domain, on the vendor’s infrastructure, and the bank pays per contract processed. It works. But the bill arrives in three places.

The switching cost stops being the code

It becomes people. Whoever processes your pipeline knows your operation better than you do — and that is a dependency no system migration solves.

It only fits above a certain contract size

And it is precisely below that size that the market is growing: FIDC managers, SCDs (direct credit companies), securitizers, credit unions, banking correspondents. Whoever lacks the volume for that contract is left with no option — and that is the desk we were built for.

The channel is a portal

The pipeline expects the customer to come in, log in and fill out forms. Here the invoice comes in through the conversation — and the agent answering is the same engine that decides, with the same data.

This is a statement about a contracting model, not about a vendor: the platform-plus-BPO model is public and has more than ten years on the market in Brazil.

Regulation and red tape

The tedious part is exactly what sets you apart.

Registration, payroll annotation, approval limits, deadlines on a clock, checking the underlying receivables. Nobody wants to build this — and now the regulation requires it. That is where an operation is won or lost.

The regulation works in your favor

Part of this is no longer a differentiator. It became a duty.

CVM 175 Annex II · art. 2, XVII and art. 36

Checking the underlying receivables is the manager’s duty.

Origin
Existence
Enforceability

One by one or by sampling — and in that case with a consistent, verifiable statistical model whose rules are made public on the fund class’s page.

And outsourcing is no way out: whoever hires a provider to check the receivables takes on the duty to oversee that work (art. 36, §5). Oversight without evidence doesn’t exist.

CMN 4,966 since Jan 1, 2025

Expected credit loss, estimated before any default.

Stage 112-month loss
Stage 2lifetime loss of the operation
Stage 3credit-impaired

If provisioning tracks risk, mitigating risk becomes a number on the balance sheet — and provisions are capital.

And this one has a date

The electronic duplicata will force a new step into your flow.

Jul · 2026Assisted production. Voluntary adoption.you are here
mid · 2027Mandatory for large payers.
late · 2027Mandatory for mid-size payers.
mid · 2028Mandatory for small payers.

If your process lives in artifacts, you add a step in an afternoon. If it doesn’t, you join the queue on someone else’s roadmap — along with all their other clients, who have the same date on the calendar.

Legal basis for the electronic duplicata (a Brazilian trade receivable instrument): Law 13,775/2018, BCB Res. 339/2023, BCB Res. 540/2025, CMN Res. 4,815/2020 and Decree 9,769/2019 — registrars authorized today: B3, CERC and Núclea; the Central Bank (BACEN) regulates and supervises, it does not register. Receivables verification is in Regulatory Annex II of CVM Res. 175, with consolidated guidance in Circular Letter 8/2025. Expected credit loss is in CMN Res. 4,966/2021.

The big losses didn’t come from defaults.

They came from fraud, misappropriation of funds and operational failure. That isn’t a score you can buy: it’s process.

The real risk

The most significant historical losses in this market did not come from portfolios that defaulted more than expected, but from fraud, misappropriation of funds and operational failures.

Jean-Pierre Cote Gil, partner and portfolio manager at Vinland Capital, to InfoMoney (translated)

That isn’t a score you can buy. It’s process — and it is the only category of loss that no data vendor solves for you.

Does it exist?

The backing is checked against the document, not against what someone typed.

Does it belong to whoever claims it?

The payer is linked by tax ID, not by name — “Unimed” and “Unimed São Paulo” are different companies.

How much does this payer usually pay?

That payer’s history goes in before the discount, not after.

And who signs when it exceeds the approval limit?

A person, with the dossier in view. In June 2026 the Financial Stability Board opened a consultation on AI in finance, with attention to agents that act without supervision — here the supervision sits where the decision hurts.

The size of the table

There’s no shortage of money in the receivables market. What’s short is operations that are ready to run.

R$ 852.7 bn

net assets of Brazilian FIDCs

ANBIMA · June 2026

3,800

FIDC classes — up 26.7% in twelve months

ANBIMA · Dec/24 to Nov/25

426

fund managers — up 14.5% over the same period

ANBIMA · Dec/24 to Nov/25

Notice the gap between the last two: fund classes are growing almost twice as fast as managers. Existing firms are launching more and more different operations — and that is where the math gets tight.

The cases

Eleven operations. Eleven risks. The same backbone.

One is in production. The others share the same backbone — what each one adds is the registration rail for its collateral.

Where this already fits

Find yours. What changes from one to the next is always the data that decides — never the process around it.

OS invoicereference

who pays · the health insurer

Doctors’ invoices against an OS (a nonprofit that runs public health units) or an insurer. The risk is the glosa (disallowance): the payer pays less than billed.

Marketplace payouts

who pays · the delivery or e-commerce platform

Cancellations, chargebacks and the amount the platform holds back.

Recurring contract

who pays · the corporate customer on the contract

Churn and SLA disputes — the receivable vanishes if the contract dies.

Trade duplicata

who pays · the buying manufacturer

Delivery performance and concentration in a single payer.

Card receivables

who pays · the card acquirer

Settlement-account locks and competing claims on the same receivables.

Freight and CT-e

who pays · the shipper

Proof of delivery and the shipper’s payment terms.

Distributed energy

who pays · the power plant’s consumer

Lower-than-forecast generation and unpaid bills.

Rent

who pays · the tenant

Default and termination before the lease ends.

Payroll loans

who pays · the employer, through payroll

Payroll registration and margin that has already been used.

Agribusiness CPR

who pays · the crop buyer

Yield, price and weather — three risks, not one.

Court-ordered debt (precatório)

who pays · the government entity

Timing and payment order, not ability to pay.

Each row answers the same five questions. Read it vertically: what changes is always the data that decides — never the process around it.

For those who want to compare

The first seven, side by side.

OperationWho assignsWho paysWhat drives the riskThe data that decidesWhat changes in the process
OS invoiceDoctor, clinicThe health insurerDisallowance: the insurer pays less than the invoiceHistorical disallowance coverage by insurer and procedureIt’s the case that is live today
DuplicataSupplier to a manufacturerThe buying manufacturerDelivery performance and concentration per payerThe payer’s history and the record at the registrarRegistry check before assignment and a limit per payer
Card receivablesThe merchantThe card acquirerSettlement-account locks and competing claimsRegistered receivables schedule and contract effectsThe collateral is the schedule; settlement becomes automatic
RentThe landlordThe tenantDefault and lease terminationLease history and the guarantee — guarantor, insuranceRecurring backing: recurrence instead of a single instrument
PrecatórioThe plaintiff owed the judgmentThe government entityTiming and payment orderStage of the case and which entity owesTiming becomes duration risk, not credit risk
Agribusiness CPRThe farmerThe crop buyerYield, price and weatherAcreage, harvest history and futures pricesSeasonal external data; physical collateral
Payroll loansThe employeeThe employerPayroll registration and available marginMargin and employment, not scorePayroll registration becomes a mandatory step in the flow

The base already runs end to end

Origination, decisioning, approval limits, disbursement and settlement — with real money, from WhatsApp to Pix (Brazil’s instant payment system). It’s what the demo runs, no login.

Every segment inherits that base

Backing, dossier, decision, disbursement and settlement don’t change from one operation to the next. What changes is the data that decides — and that is your competitive edge.

The registration rail is scope, and we measure it

Registrar, card receivables schedule, notary, payroll registration: each segment has its own. We tell you which one and how much before the proposal — never after the contract.

How we work: the base is the same for all eleven, and each segment’s registration rail comes in as measured scope — registrar, card schedule, notary or payroll registration, depending on the collateral. We’d rather tell you which one and how much before the proposal than tick a box the contract can’t back up.

Next step

What’s your operation?

Bring the one you already have, or the one you want to launch. In thirty minutes we can tell what can be reused and what needs to be connected.

How we deliver

In days, because it’s assembly.

What takes time in a credit project was never the screens: it’s the foundation. And the foundation is already standing, in production.

How much work this takes

Three columns, and your whole operation fits in them.

already built
you configure
we integrate
Origination and intakeThe invoice comes in over WhatsApp or a portal
Deal dossierBacking, history and proposal in one place
Decision ruleThe bands and cutoffs are your policy
Approval limit and pricingHow much it approves on its own, and what it costs
Proposal textWhat your customer reads
Disbursement via PixAuthorized, with an audit trail
Automatic settlementOnce paid, the receivable settles
Collateral registrationRegistrar, notary, payroll registration
Your ERP and your credit bureauWhat you already have and won’t replace
What this means
It exists and runs in production. You don’t pay to build it again.
Your credit team changes it on screen, with no ticket and no sprint.
It’s the scope we measure before the proposal — never a box already ticked.

Seven of the nine rows already exist. That’s why the first operation ships in days — and why the second ships faster than the first.

And they combine

Block plus block becomes a process — with no code in between.

Decision engine+Operation dataarrow_forwardThe approval limit you designed, deciding with that payer’s history.
Account and Pix+Eventsarrow_forwardThe insurer pays and the receivable settles on its own — nobody checks bank statements.
Conversation+Decision enginearrow_forwardThe invoice comes in over WhatsApp and the proposal goes back through the same channel.
Identity+Dossierarrow_forwardThe desk sees the whole operation; the seller sees only theirs. The same data, two views.

The distinction that matters to a risk committee

This is not “build an app from a prompt”.

what’s being sold out there

Code generated on the spot

Gartner projects that prompt-to-app approaches adopted by people outside IT will increase software defects at scale through 2028, with a crisis in quality and reliability.

Every change is new software, and new software has new defects.

what happens here

Configuring tested blocks

Changing an approval band doesn’t generate software: it changes a value in a block that has already been tested and is already in production — in your staging environment first.

AI comes in during assembly, done by people who know what they are assembling.

What we measure first

“In days” has fine print, and it’s short.

What’s fast

The backbone: origination, dossier, decision, approval limit, disbursement and settlement. It already exists.

What’s variable

Integration with what you already have — the fund’s ERP, your credit bureau, your core system, the registrar for your collateral.

That’s why we measure first

The thirty-minute conversation separates what can be reused from what needs to be connected, before any proposal.

Continuity

The final product is yours.

Investing in the platform that runs your business can’t become a continuity risk. The answer isn’t a promise — it’s three concrete things.

It’s the right question, and it always comes up: what if I become dependent? The answer isn’t a promise — it’s three concrete things.

The final product is yours

Your operation doesn’t live in our code: it lives in artifacts — the flow, the decision table, the approval limit, the pricing, the proposal text, the data.

An artifact is something you open, read and change.

Your team learns to evolve it

We train your credit and operations team, not your IT. The people who change the rule are the people who understand the risk.

That’s on purpose: technical teams want to write code, and code is exactly what this approach avoids.

The exit is in the contract

Escrow, a perpetual license to the delivered version, or assignment triggered by an event: the term that closes this risk is chosen with you, and it goes into the contract.

It isn’t fine print — it’s the clause that gets the committee to approve.

Where that leaves Catalisa

Two questions separate the vendors of credit software.

Who can change the process? And who owns the product in the end? Almost every vendor answers one well and the other badly.

only the vendor changes it the business team changes it vendor’s product client’s product Custom-built project The product is yours, but every change is a sprint — and depends on technical people. Catalisa Ready foundation, your product, and the process in artifacts the credit team edits. Quick to start, cheap to change. Off-the-shelf system Quick to buy, and your process becomes their process. Infrastructure parts Great blocks — but whoever assembles the process is still your IT.

This quadrant is ours, not Gartner’s. Gartner doesn’t evaluate Catalisa, and saying otherwise would be inventing a credential. The axes are our reading of the market — and the category Gartner would put us in is credit origination.

If you want to go further

White-label, running on your infrastructure.

It’s not the main pitch, and that’s on purpose: for most clients, team autonomy and a contractual exit already solve it. But when the committee requires data and process to stay in-house, that door exists.

What it is

The platform runs in your environment, with your brand and your domain. The operation’s data never leaves it.

Who runs it

Your team, trained by us. Updates come by version, in your window — not when it suits us.

What gets written down first

Environment requirements, who answers for incidents, and how a new version goes in. Without that, white-label is just a promise.

Where we don’t go

What Catalisa is not.

  • close
    Core bankingWe don’t compete with the ledger. We sit next to it.
  • close
    Registrar or credit bureauWe consume both. The bureau says whether the CNPJ pays; we are the process that uses it.
  • close
    Fund manager or fiduciary administratorWe don’t manage funds or keep their books. Your FIDC ERP stays where it is.
  • close
    A promise of approval, rate or returnThe desk decides. The platform organizes the decision and keeps the reason why.

Next step

Start with the hard question.

Bring the committee. Continuity questions are the ones we answer best, because they’re the ones we’ve already been asked.

Credit desk

Restricted to the credit review team.

See it working

The invoices go into the system and the analysis really runs — it’s not a recording.

account_circle
Receivables online

What happens behind the scenes

Send the invoices on the left to start.

hourglass_top Loading your invoices and your proposal…