The LLM Agents Hackathon, organized by Berkeley RDI in conjunction with LLM Agents MOOC, has come to an end. The second place went to a team of two employees from IT company SoftServe with a project to find the next generation of missing persons.
The LLM Agents Hackathon, organized by Berkeley RDI in conjunction with LLM Agents MOOC, has come to an end. The second place went to a team of two employees from IT company SoftServe with a project to find the next generation of missing persons.
The LLM Agents Hackathon, sponsored by Berkeley Research Institute, lasted two months, where each developer could learn how to use LLM Agent technology to create innovative applications (Applications Track).
The event included five tracks focused on the development of Generative AI agents:
The jury included professors from UC Berkeley, researchers from Google DeepMind, OpenAI, and Meta AI. The winners will receive prizes from these companies, with a total prize pool of up to $200,000.
The hackathon was attended by 3,000 representatives from 127 countries, over 1,100 universities, and over 800 companies. The most represented countries were the USA, India, and China.
Among the participants were representatives of top technology companies, including Amazon, Microsoft, Samsung, and Salesforce.
The second place winner of the hackathon was the ThreadFinders team, consisting of two SoftServe employees: Maryna Bautina and Nataliya Polyakovska. The team members created a system of interconnected agents based on Google Cloud Platform to search for missing people. The developers from ThreadFinders have been best friends for six years, but have only met in person five times — all at conferences — because we constantly live in different places.
This system creates a single hub for volunteers searching for missing persons, divides the map into search zones, tracks updates on where search operations have already been conducted, integrates with news (for example, if there is a traffic accident or other event nearby that may be related to a missing person), and generates alerts to organize searches at critical points.
The next generation of missing persons searches is being carried out thanks to:
