Work the right cases first
Your fraud alerts, AML cases, disputes and collections pile up faster than your team can clear them. The order your team works them in costs you money. We work out what today's order costs you on your own history and give you a better rule that you can explain to a regulator.
What your control team decides every day
-
Which cases go first
Working the biggest exposure first feels safe. With the hours you have, it is usually not the best choice, and we show you how much it costs.
-
Which cases are worth a look
A review costs analyst hours whether it finds anything or not. Only some cases have enough expected loss to be worth those hours.
-
Where the threshold sits
If a rule catches more fraud, it also stops more good customers. Both cost money, and the threshold is usually set once and left for years.
-
How many analysts you need
Work does not arrive evenly. If you staff for the average, you get a backlog. If you staff for the peak, you pay people to wait. We work out how many you need and when.
Try it on your own numbers
Type in your cases, what each one exposes you to and how long it takes, then set how many hours your team has. The tool picks the best set and shows what working biggest-first costs you.
It runs in your browser, so what you type stays on your machine.
Where it applies
- Fraud and dispute queues at payment companies.
- AML alerts and credit exceptions at banks and lenders.
- Collections teams with more accounts than calls.
How we work with you
-
1. Talk
Pick the queue that hurts most
Tell Jiri or Ivo about it. Talking to us costs nothing. If you want proof first, a short audit shows what your current rule costs.
-
2. Proof
Test it against your baseline
We run the new rule on your own history next to today's rule. How we test it.
-
3. Build
Build what works
If it does better, we build it into a system you own. How we work.
What we need to see
The queue, past cases with how they ended and how long they took, and how many hours your team has. It stays confidential, and Jiri or Ivo goes through it with you.
Rules that must never break
Some rules can never go past an exposure cap or a contract limit. Tests check the cases someone thought of. A proof covers all of them, and it costs less than a breach.
Rainbow is our theorem prover for Clojure. We use it to prove those rules hold.
What you get
- A system built around the controls you already have.
- A rule you can write down, with the reason for every case on record.
- A machine-checked proof for the limits that must never be crossed.
Where we have worked
Ten years in, about 50 people across around 20 countries. We built financial and transactional systems for these companies.
-
Zimpler
Fintech and mobile engineering.
-
Guaranteed Rate
Software engineering engagement.
-
Cicada Wallet
Blockchain wallet engineering and product design.
-
Starcity
A payment microservice pulled out of a monolith.
Bring us one queue
Pick one, like your fraud queue or AML review list. Jiri or Ivo will tell you what your current rule could be costing you. Talking to us costs nothing.