With the emergence of ChatGPT and other large language models (LLMs), the demand for vector databases has skyrocketed. However, most companies still rely on fragmented infrastructure and legacy tools that aren’t built to handle AI-native search workloads. TopK solves this problem with a “true hybrid retrieval” approach that combines vector search, keyword matching, and filtering in a unified engine. TopK’s purpose-built query engine gives teams deep control over retrieval and ranking — enabling fast and high-quality search in large-scale production environments. The company has now secured $5.5 million in seed funding from Earlybird, KAYA, Irregular Expressions, and several angel investors. The funds will be used to expand the team and build out TopK’s enterprise offering to support web-scale search use cases.
„The database market is undergoing a fundamental transformation. The boom in AI-native products is pushing the limits of traditional search systems, which no longer meet current demands. Our platform addresses this shift in the market and provides a solution for developers and AI agents to unify structured and unstructured retrieval at scale.” says Marek Galovič, CEO and co-founder of TopK.
“We are changing the way data is searched and building a platform tailored for development teams that work with large volumes of data and demand high performance, flexibility, and cost efficiency. Our goal is to become the unified query layer that powers search for the world’s most advanced apps. That's why we welcome design partnerships that challenge us with technically demanding problems and help us push the product even further," adds Jerguš Lejko, CTO and co-founder of TopK.
The cloud-native platform combines vector search, keyword matching, filtering, and flexible ranking in a single system. This allows teams to fine-tune results for their specific domain and use case, for example, in e-commerce, finance, healthcare, or law. In the future, TopK plans to expand its platform’s capabilities with natural language understanding and support for unstructured and multi-modal data. This will allow the startup to move beyond a traditional search database and deliver truly intelligent search.
AI transforms how we search and understand data
TopK was founded in 2024 by Marek Galovič and Jerguš Lejko, who have known each other since high school. After years of working on large-scale systems and AI tools, they reunited again to reimagine search from the ground up. Their goal was to provide teams with a unified platform for relevant and intelligent search across all their data – without the need to connect multiple databases, APIs, or ranking systems.
„Companies are facing the limitations of current search systems. TopK brings a fundamental shift in how infrastructure is built, along with a deep understanding of the problems that development teams face every day. That’s why we support their vision to create a search system for the AI era,” says Andre Retterath, General Partner at Earlybird.
„TopK has built the core of their database to deliver the fastest and most cost-effective hybrid search solution on the market – all with a small but exceptionally capable team. The investment will allow them to accelerate the development, deliver the enterprise-grade features, and succeed on a global scale,” says Karel Zheng from KAYA.