Over the course of ten days, participants developed solutions for real business challenges using generative technologies and other ML models. Over 300 Master’s students at AI Talent Hub, AI enthusiasts, and experts took part in the event.
The hackathon featured 13 cases offered by the companies X5Tech, Raft, Napoleon IT, Ugra Research Institute of Information Technologies, Nornickel, and Severstal’s Laboratory of Measuring Systems. Among the tasks were: development of an AI service for generating web interfaces, monitoring of toxic content in AI products, research into the vulnerabilities of software and network attacks, development of a Russian-Mansi translation software, forecasting of demand for a retail company, development of a tech support chatbot, and so on.
Nornickel provided a task on the development of a system able to process and analyze regulatory documents associated with the construction of objects. The resulting system needed to be able to not just search for and offer information, but also answer complex questions related to regulations and standards based on the analysis of documents it was given. The best prototype was developed by the team Улыбка Мона Лизы (“Mona Lisa’s Smile”).
The team Ailab presented a solution for monitoring toxicity in artificial intelligence products for the AI developer Raft. The solution was designed for industrial applications; its architecture makes it possible to connect analyser models by other teams via a common interface.
The team Нас пятеро (“Five of Us”) created a chatbot for X5Tech that automates the first line of tech support.
“The participants showed their ability to delve deep into working on a case’s particulars and did their best to optimize the associated business metrics, which was going even further than what the task implied. This demonstrated the Master’s students’ skills as not only programmers, but also capable product specialists. The winning teams’ projects demonstrated high precision in closed benchmarks. I hope we will continue to collaborate with the Master’s program and nurture new talents,” comments Michil Egorov, the head of X5Tech’s development team.
AI Product Hack 2024. Photo courtesy of the organizers.
The teams Улиточки (“The Snails”) and ЖигулИИ (a play on “Zhiguli” and “AI” – Ed.) developed a Telegram bot for Наполеон.Отзывы (Napoleon.Reviews – Rus.), which contacts a client, tells them about the product, presents a demo, answers questions, as well as prepares them for communication with the sales department.
The team Baseline Solutions did best with the case by Severstal’s Laboratory of Measuring Systems. The participants developed a computer vision-based algorithm for precise identification of an object’s 3D geometry, which can help to considerably increase quality control and facilitate development at industrial enterprises.
“I’d like to note that even professionals with years of experience would’ve been astonished with the approaches proposed by ITMO students. The participants showed solutions with a high potential for further development within the R&D project that we plan to work on,” says Aleksandr Nikitin, a project manager at Severstal’s Laboratory of Measuring Systems.
AI Product Hack 2024. Photo courtesy of the organizers.
According to the organizers, the hackathon has not only become a platform for finding new solutions, but has also helped companies test promising technologies that have the potential to transform their businesses in the future.
“We’re seeing a huge demand for AI-based solutions, more language models and generative approaches, but also a catastrophic lack of specialists and expertise in new technologies – both in-house and outsourced. As a major Master’s program in AI and a hub on LLM technologies, we conduct AI Product Hack precisely with the purpose of giving companies the opportunity to test hypotheses and introduce innovative solutions with the participation of specialists who develop breakthrough products based on frontier technologies. This is a unique chance for companies to receive invaluable experience of working with the newest AI methods even before those enter the market,” underlines Pavel Podkorytov, the CEO of Napoleon IT and co-founder of AI Talent Hub.
AI Product Hack 2024. Photo courtesy of the organizers.
Aleksandr Ivachev, the head of the team Design Suite NLMK, believes that participating in AI Solution Hack gave them an opportunity to not only see interesting new solutions, but also to estimate how quickly and efficiently it is possible to solve tasks with the help of artificial intelligence.
“In the past 4 years, our team has done a lot in terms of creating and developing Design Suite NLMK, and we continue to improve it. As someone who contributed a case for the hackathon, I was really impressed with the teams’ expertise, and I’m already discussing the opportunities for further cooperation with my colleagues,” he notes.
