DATFID
Entry last updated on April 1, 2026

DATFID transforms black-box AI forecasting into transparent, traceable decisions.

Born from over ten years of research in quantitative analysis, econometrics and software development, our software delivers instantly interpretable forecasts for entire portfolios, without training, tuning or GPU infrastructure.

DATFID acts as a forecasting layer that integrates seamlessly into existing systems without complex adjustments or system changes. By considering dependencies between products, locations, or other entities over time, DATFID improves the quality of results by up to 20% compared to conventional methods and reduces the overall analysis effort by more than 50%. Companies not only understand what will happen, but also why: DATFID makes cause-and-effect relationships visible, builds confidence in the results, and enables hedge fund-level predictive analytics without the need to build a PhD team.

Show English Description

Facts
 
founders
 
Contact
 

2

Founders

2026

Founding year

Startup stage

Stage

B2B

Business model

Founder Igor Schapiro, CEO, M.Sc. Computer Science (TUM), and Dr. Valerii Dashuk, CTO, Ph.D. Quant. Econometrics (LMU)

Company location

DATFID
Rosenstrasse 2
82205 Gilching

You might also be interested in

The EPO headquarters in Munich

News

Munich leaves everyone else in the dust: Nowhere else in Europe are there more patents.

Munich is further consolidating its role as a central innovation hub in Europe: No other city unites so many patent applications, research institutions, and technology-driven companies in such a small area…

Tubules

News

Tubules purchased from American pharmaceutical giants

Munich-based biotech unicorn Tubulis is being sold to US corporation Gilead for a billion-dollar sum.

News

Milk without cows – progress with patience

Munich-based startup Precision Labs combines high-quality milk protein with plant-based ingredients to create a milk alternative that is as good as possible in terms of taste, function, and nutritional value…