A look back at the first Constructor University Causal AI Datathon — three days of causal reasoning, industry collaboration, and hands-on optimization on real manufacturing data.

The Constructor University Causal AI Datathon 2026 brought together 50 registered teams at Constructor University Bremen for inverse causal optimization on an industry dataset — not a prediction contest, but a challenge in identification and intervention design, with Insytes GmbH as industry partner.

Overview

Event summary

The problem

Teams received a real-world injection molding manufacturing dataset and a causal DAG. The task was inverse causal optimization: given a target reduction in scrap rate (the KPI), identify which controllable process variables to intervene on — and by how much. This was not a data-science or prediction contest; it required causal identification and working backwards from a desired outcome to optimal interventions.

Program highlights

Day 1 — April 17

• Opening remarks by Prof. Keivan Mallahi-Karai
• Keynote by Omar Elshinawy (former Math Society president)
• Industry talk: Introduction to Causal AI by Varun Patel (Co-CEO, Insytes GmbH)
• Panel: From Proof to Product: How Mathematical Reasoning Shapes Industry — Omar Elshinawy, Karim Abualrish (Co-CEO, Insytes GmbH), Prof. Keivan Mallahi-Karai

Days 1–3: Open research and development period for all teams.

Day 3 — April 19

• 12:00 — Paper submission deadline
• 15:30 — Team presentations (6 minutes per team + Q&A)
• Judging and deliberation
• Awards ceremony and closing dinner

Judging

Day 1 — April 17: Opening, talks & challenge launch

Opening by Prof. Keivan Mallahi-Karai, keynote by Omar Elshinawy, an industry talk on Causal AI by Varun Patel (Insytes), and the panel From Proof to Product — then teams entered the open R&D period for the inverse optimization challenge.

Insytes co-founders & Omar Elshinawy

Day 1 brought together Karim Abualrish and Varun Patel (co-founders of Insytes GmbH) with Omar Elshinawy (former Math Society president) for the keynote and the From Proof to Product panel — connecting mathematical reasoning to industry — with Ahmed Maghri (Math Society president) helping anchor the opening day.

Day 2 — April 18: Research & development

Part of the three-day open R&D window: teams continued inverse causal optimization — identifying interventions on controllable variables to hit the scrap-rate target — with support from peers and mentors as needed.

Day 3 — April 19: Submissions, presentations & awards

12:00 — paper deadline · 15:30 — team presentations (6 min + Q&A) · judging · awards ceremony and closing dinner.

Presentation highlights

Results

1st place — Team PS
Seyedpouria Seyedi

2nd place — Team DonBathon
Yaroslav Molybog, Ivan Balashov

3rd place — Team OMEGA
Amaan Shaikh, Vamshi Jinka

Winner photos

First, second, and third place — podium moments for each team.

1st place — Team PS
Seyedpouria Seyedi

2nd place — Team DonBathon
Yaroslav Molybog, Ivan Balashov

3rd place — Team OMEGA
Amaan Shaikh, Vamshi Jinka

Winners

Judges & speakers

Sponsors & Partners

The Datathon 2026 is organised with the support of Constructor University as the host institution and Insytes GmbH as the industry partner. Insytes GmbH contributes the challenge problem, dataset, mentorship, and judging — bringing real-world industry expertise directly into the competition.