The Showcases
of Real-world
AI Impact

The COVID-19 pandemic highlighted the need for powerful tools to simulate the spread of disease as closely as possible to reality. The solution needed to model contact networks between individual agents and be ready for simulation of other population diseases in the future.

We set three main goals:

1. Realistic Simulation: Base predictions on a synthetic population reflecting detailed demographic, social, and economic characteristics of various regions.
2. Simple Deployment: Create a system usable in computing centers or the cloud despite computationally demanding tasks.
3. User-friendly Interface: Enable easy manipulation of data and definition of parameters.

Our client Dormer Pramet, the global manufacturer and supplier of tools for the metal cutting industry, faced a quality control problem. Their complex manufacturing process relied heavily on manual inspections, leaving the risk of undetected defects until a final inspection or even worse – by the customer.

Tens of thousands of inserts with various features made thorough inspection challenging. Tiny defects, like microscopic cracks invisible to the naked eye, required specialized equipment. The lack of detailed defect data also hindered analysis and improvement.

Our client was looking for an AI-based visual inspection solution that would increase efficiency, reduce costs and deliver flawless products.

The Institute of Molecular and Translational Medicine (IMTM) aimed to create a comprehensive database of the Czech genome as part of the ENIGMA project. This involved processing 700 sequenced genomes from Czech donors to compute population-specific characteristics of genetic variants.

However, they faced significant challenges:

• Data Volume: Traditional methods could not handle the massive amount of data.
• Storage and Accessibility: Existing solutions could not efficiently store, query, and retrieve such vast genomic data.
• Resource Constraints: Developing the database in-house would require substantial computational resources and personnel.

Shift planning in healthcare is not just an administrative task. In the AKESO healthcare group, it means coordinating operations across 21 entities, more than 2,000 employees, and approximately 100 departments, centers, and outpatient units.

Every month, more than 30,000 shifts and duties need to be planned. The schedule must take into account staff availability, qualifications, legal requirements, operational needs, and employee preferences.

DPD needed to verify whether the MyIO platform could accurately forecast parcel volumes across its logistics network. The goal of the PoC was to retrospectively forecast parcel volumes for the first quarter of 2023 and validate forecast accuracy at different levels of detail.

The project worked with approximately 45 million parcels per year, and the forecast needed to cover the hierarchical structure of the logistics network — from the global level through countries and regions down to individual postcode areas.

Masoprofit is a supplier of equipment and technologies for food service, butcheries, and food production. The company manages a portfolio of more than 20,000 products, whose demand changes significantly over time and is influenced by seasonality, marketing campaigns, and discounts.

Inventory planning was primarily based on historical data and manual decision-making by buyers, which led to stockouts and inefficient inventory levels.

The pharmaceutical company Zentiva needed to significantly accelerate the process of calculating manufacturing costs for medicines, known as Cost of Goods Sold — COGS. This calculation is crucial when developing new formulations and modifying manufacturing processes.

COGS calculation is a highly complex process involving a large number of parameters, including raw materials, manufacturing steps, machine utilization, labor, and packaging. The traditional approach required manual analysis of these factors and was time-consuming.