Uzbekistan’s Cerberus Wins Road to TechCrunch Regional Final

Uzbek startup Cerberus has won the regional final of Road to TechCrunch Startup Battlefield 2026, earning a ticket to San Francisco. For Uzbekistan’s young technology industry, the result is another sign of its push beyond the domestic market. For Cerberus, the challenge is now to convince international investors and clients that an AI tool developed in Tashkent to find vulnerabilities can compete in the global cybersecurity market. Cerberus took first place in the regional final on August 12. WeGlobal AI finished second, followed by LOOQ. Twenty-two startups reached the final after being selected from 726 applications from 39 countries. Uzbekistan was represented by six projects. The three winners will represent Central Eurasia at Startup Battlefield 200 in San Francisco. TechCrunch Disrupt 2026 will run from October 13 to 15. Twenty of the 200 companies will be selected to pitch on the main stage, with five advancing to the final round to compete for the $100,000 equity-free grand prize and the Disrupt Cup. The three regional winners will share a $100,000 investment pool, with Cerberus receiving $50,000. Competition rules state that each investment will be made on standard market terms in exchange for equity. Astana Hub Ventures and IT Park Ventures are providing the funding in partnership with Silkroad Innovation Hub. According to the organizers, the winners will also receive a combined $100,000 in OpenAI API credits. Cerberus founder and CEO Aziz Akhmedkhodjaev told The Times of Central Asia that two factors were behind the company’s success: the product and the team’s ability to explain the problem it solves. “I would say there are two main things here. The first, of course, is the product. Everyone understood that we had created something the market really needs,” he said. “I think that was the key to winning. I didn’t spend a single second of the pitch on anything other than clearly explaining the problem, how the solution works, and its potential.” Cerberus describes its product as an AI-powered cybersecurity system capable of finding vulnerabilities in corporate applications and IT infrastructure. Demand for such tools is growing alongside the spread of generative AI: companies and individual developers can write software faster, but the volume of code that needs to be checked for errors and potential attack vectors is also increasing. “Literally everyone is writing code with artificial intelligence now. Vulnerabilities, coding errors, and other weaknesses are appearing much faster,” Akhmedkhodjaev said. The company says Cerberus allows users to give its AI agent access to an application or IT infrastructure and task it with finding vulnerabilities. The agent can also test authentication and perform other security checks. According to Akhmedkhodjaev, however, using a general-purpose AI model for such work can itself create risks. “If you give Claude or another AI access to a vulnerability in a real bank, even the smallest mistake can lead to serious financial losses,” he said. The company says Cerberus is designed to restrict what its AI agent can do after identifying a weakness. “In our project, there is a...