Knowledge discovery & ML prototypes
Discover patterns and rules in complex data and develop an ML prototype.
Explore the service →
Knowledge Discovery · Decision Support
Turn complex data into informed decisions.
IAP develops custom AI and explainable decision support for complex business problems. We analyse your data, develop the methods needed and implement them in software, from the research question to a deployed MVP and integration into your workflows.
IAP since 2000 · Scientific and technical leadership: Prof. Dr. habil. Michael Thrun
More than 25 years of company history
Original method development
Research through usable applications
Where IAP creates value
AI for problems that require original methods.
Your data is complex, a standard model is insufficient or your experts need an understandable basis for decisions. IAP combines scientific depth with the development of a practically usable solution.
Decision support & explainable AI
Decision support that explains results and flags uncertain cases.
Explore the service →
Custom AI & method development
Develop original AI methods and deliver them as a prototype, MVP or integrated application.
Explore the service →From problem to result
Three examples of project work
Commissioned development · MVP
Forecast demand and support supply-chain decisions
Future product demand needed to inform planning in a supply chain with an automated warehouse.
IAP’s contribution: IAP developed a custom forecasting AI and implemented it in a decision support MVP designed for supply-chain planning in an automated warehouse.
Project details →
Commissioned development · delivered solution
Condition-based maintenance from operating data
Train operating data was to provide indications of conditions relevant to maintenance. Changes caused by varying loads and operating conditions needed to be distinguished from possible component faults.
IAP’s contribution: IAP developed an anomaly-detection approach with DB Cargo’s predictive maintenance department. Maintenance experts’ hypotheses were investigated using historical measurement and maintenance data. Models describe relationships between sensor readings, and deviations from those relationships help characterise unusual conditions.
Project details →
Commissioned prototype based on scientific research
FlowXAI: explainable support for lymphoma classification
Interpreting flow-cytometry data requires expertise. Rare entities, different measurement conditions and limited training data complicate automated classification.
IAP’s contribution: IAP used research findings from the PLAIT author group to develop an explainable AI prototype for Beckman Coulter. FlowXAI analyses the samples and returns classifications, explanations and reliability assessments.
Project details →
What customers value about working with IAP
Jürgen Swora
SW-MOTECH
The custom-made forecasting AI developed by IAP-GmbH provides important insights into future demands for our products that allows us to reduce overall supply chain costs. The acquisition of the competitor off-the-shelf software would have been ten-times more costly with no better forecasts.
Role at the time of the statement
General Manager
Original statement · English
Thomas Pohl
DB Cargo
IAP-GmbH provided a solution for a problem where the conventional approaches fell short.
Their decision support system for predictive maintenance provided us with valuable insights into our data and now enables automated predictions of machine failures.
These prediction will play a vital role in our decision-making process, allowing us to minimize downtime, reduce costs and ensure smoother operations.
Role at the time of the statement
Head of Predictive Maintenance, DB Cargo AG
Original statement · English
IAP displays the 2021 and 2025 BSFZ seals for certified research and development activity.
Local AI service from IAP
Local AI for confidential business data.
IAP configures Apple hardware, a locally running AI model and the agreed tools for your task. This enables confidential company data to be used for analysis, forecasting and decision support within your infrastructure. The scope is set out in an individual quotation.
View scope and indicative price
Approved company data
↓
Local language model: questions and explanations
↕
Analytical tools: computed results
↓
Review and decision by your experts
Example of a local workflow
Scientific foundations
Research that can be examined and built upon
Thrun et al. · PLOS Medicine · 2026
FlowXAI: explainable AI with case-level reliability assessment
Structure discovery, explicit classification logic and assessment of individual predictions are combined for lymphoma classification.
Stier, Gehlert & Thrun · Multiresolution Forecasting for Industrial Applications · Processes · 2021
Wavelet forecasting for industrial applications
The paper investigates wavelet-based multiresolution forecasting for industrial time series. It combines features at different time scales with systematic comparison of forecasting methods.
Which decision do you want to improve?
In a free, non-binding 20-minute introductory discussion, we clarify your objective, starting point and a useful next step.