is the AI company building information infrastructures

for the granular credit market


Algoritmica partners with data providers to deliver high-quality, integrated datasets to banks, investors and insurance companies

Why Algoritmica?

Our AI-based software quickly and accurately assesses default events and credit losses, up to 50% quicker and 10% more accurate than traditional applications.

We help investors and institutions reduce their credit losses and make better lending decisions.

Algoritmica is the company behind deeploans, an AI-based data processing software that feeds market participant’s analytics and workflows with unique data points


Efficient deal flow

Find deal-useful credit drift by generating custom scenarios


Quantify loan losses up to 36 months in advance by leveraging pre-trained Temporal Fusion Transformer models


Unique data

High quality loan-level data coming from trusted data providers

Built-in compliance

Data sharing infrastructure and legal framework for a secure transmission and dissemination of information

We are building an open and transparent infrastructure for aggregating and analysing loan-level data.
Connect with an expert to learn more.

Frequently asked questions

deeploans is a software and have at its core machine learning-based algorithms to accurately detect data quality issues related to credit events and credit losses in the loan-level data

deeploans trains with loan-level data from over 250 financial institutions in Europe. 

deeploans analyses numerous data points to identify data quality issues related to the borrower, their business, industry, and collateral, such as:

  • Business Confidence indexes & Real Estate Price indexes
  • Arrears benchmarks on multiple levels (NACE, Postcode, firm size, etc.)
  • Profitability Ratios (Working Capital, Net Profit Margin, etc.)
  • Liquidity Ratios (Free Cashflow Ratio, Absolute Liquidity, etc.)
  • Sales trends
  • Employee wages
deeploans also allows data users and providers to manage consents for accessing data.

Start by using the sandbox