What people ask before signing.
If yours isn't here, write to comercial@catalisa.app.
What is a credit decision engine?
It's the system that applies your credit policy to each application: it reads the data, evaluates the rules and returns approve, decline or review. Here the policy is a DMN decision table, versioned and published without a deploy, and every decision stores the rule and version that made it.
Do I need to replace my system?
No. The rules move out of the code and become a call; the rest of your service stays exactly as it is. If you already have the data, use just the engine.
How long does it take to get to production?
Integrated — you bring the data — one afternoon. With the pipeline configured, a few days. The industry usually takes around twelve months from concept to production; that's exactly what we're going after.
What is DMN, and why does it matter?
DMN (Decision Model and Notation) is the OMG's open standard for writing business rules as tables. It matters for two reasons: people who don't code can read the policy, and it leaves here in one piece if you ever want to take it elsewhere.
What if I want to leave later?
Your whole policy exports as DMN, an open standard. No proprietary format locks your operation in — and your decision history can be exported.
Does it work for decisions other than credit?
Yes. The engine evaluates the rules you wrote, whether for fraud, eligibility or onboarding. What's specific lives in your policy, not in the product.
Does AI decide on credit?
No. AI reads your decisions, points out patterns and explains why each one came out the way it did. What decides is the published table — same input, same output, with the version recorded.
Do you resell bureau lookups?
No. The pipeline uses your operation's contract and credentials. The vendor bills you directly for each lookup, at the price you negotiated.
Is there a minimum contract or commitment?
Neither. The Free plan is forever and paid plans are monthly, with no lock-in.
How does the Start plan's 30-day free trial work?
Thirty days with the full Start plan — 3,000 decisions a month, engine and pipeline — at no charge. The trial applies to Start; Growth and Scale are billed from day one. If you don't continue, the account goes back to Free and keeps working with 1,000 decisions a month, without losing any policy or history.
What happens when the allowance runs out?
On Free, the next call answers 402 with the amount used until the month turns over. On paid plans, anything above the allowance is billed as overage, at the plan's per-decision price.
Is there a test environment?
Yes. The test key is unlimited, never bills and can't see production data. The call is the same; only the key changes.
What about compliance with the Central Bank of Brazil?
The platform runs on our own infrastructure in Brazil — which, under CMN 4,893, means a notice within 10 days rather than a 60-day prior authorization. And every decision keeps the version and the trace that model validation asks for.
And the LGPD?
Every decision keeps its input, sources, rule and version: that's what the review required by art. 20 of the LGPD, Brazil's data protection law, needs. The criteria come from each decision's trace.