Skip to content

Algorithms make mistakes. We find them.

We provide independent reviews of algorithmic systems in banking and insurance.

To find errors, logic flaws and potential biases that are easy to miss.

So you can fix them before customers or partners are affected. 

 

We're usually brought in to look for:

  • Calculations that don't match what customers were told.
  • Commissions that don't match broker or partner agreements (paid too high or too low).
  • Biased claims fraud models that waste investigators' time and treat some customers unfairly.
  • Flawed scoring that rejects good customers or approves risky ones.

 

How our engagements work:

  • We check results against what was intended (contracts, product terms, policies), not just whether the process was followed.
  • You work directly with Yusuf, who has more than 20 years of experience reviewing systems for banks and insurers.
  • Fixed fees, usually between $50,000 and $150,000.
  • Most reviews take 2 to 6 months.
  • We're independent: we don't design or build the systems we review.

Why is this so hard to get right?

Most problems we find aren’t because teams are careless. FS data and algorithms are hard to get right.

  • Layers of legacy: errors often occur when old and new systems connect (overlap or interface).

  • Drift: a model that worked perfectly in 2020 might be unfair today.

  • Black boxes: it's not always easy to explain how a fee or decision was calculated.

 

How we review

We go beyond reading policy documents and mapping processes.

We trace the calculations, test the logic, and check the data that feeds it.

Then we tell you in simple terms what’s wrong and where to fix it.

Our approach is outlined further here.

 

We work with senior leaders

Heads of claims, fraud, credit and partnerships/distribution in insurers, banks and credit unions.

People who want to know their systems are accurate and fair.

 

Explore our articles? Start here: