UNIT.City — місце, де люди працюють... КРАЩЕ! Обирай свій простір просто зараз 👉
Олег ОнопрієнкоAI Eng
27 January 2026, 11:33
2026-01-27
AI tutor for Mriia, self-learning agent and corruption detector for NABU: $7,500 prize pool up for grabs at LAPATHON 2026
The large-scale hackathon LAPATHON 2026, which brought together developers to create innovative AI solutions for education, business, and the public sector, has ended in Ukraine. The total prize pool of the competition was $7,500 in cash and $2,000 in cloud credits.
The large-scale hackathon LAPATHON 2026, which brought together developers to create innovative AI solutions for education, business, and the public sector, has ended in Ukraine. The total prize pool of the competition was $7,500 in cash and $2,000 in cloud credits.
The hackathon was supported by leading technology companies and government institutions: mono, Genesis, Automat-IT, SoftServe, Nvidia, as well as the Mriya application and NABU. Location partners were APPS UCU and KSE.
The competition took place in three thematic tracks, each of which aimed to solve applied problems using artificial intelligence.
"Dream" Track: Educational AI
Participants in this direction worked on creating an AI tutor — a digital assistant that should provide a long-term educational effect. The task was to develop an agent that would not just give ready-made answers, but would help to understand the material.
The solution functionality included generating structured notes, creating adaptive personalized tests, and analyzing student performance. The teams used real-world data and the Lapa tool for validation.
Winners of the track (prize fund — $3000 from mono):
1st place ($1500) — Build Future
2nd place ($1000) — LLMavericks
3rd place ($500) — Initium
Genesis Track: Self-learning AI
This track, titled Tabula Rasa Agent, challenged developers to create a universal AI agent capable of continuous self-learning. The main challenge was to teach the system to gain knowledge through dialogue, resolve conflicts in data, and explain its decisions based on its learning history.
To test the effectiveness, the agents had to solve algorithmic problems in the Ukrainian programming language "Mavka".
Track winners (prize fund — $3000 from Genesis):
1st place ($1500) — Fly Me To The LUN
2nd place ($1000) — HoHoLLM
3rd place ($500) — Bugs As A Service
NABU Track: Technology against Corruption
The track from the National Anti-Corruption Bureau of Ukraine focused on automated processing of large data sets. Teams created an AI platform capable of analyzing data on individuals from state registers, identifying hidden connections and potential violations of the law.
Track winners (prize fund — $1500 from Automat-IT):
1st place ($1500) — JustSolution
2nd place — Lviv Mahjong Club
3rd place — ShocoLapky
In addition to cash prizes, the teams received support from SoftServe and Nvidia in the form of $2,000 in cloud credits for the further development of their projects.