As generative AI and large language models like ChatGPT rapidly grow, enterprises encounter new data searching and retrieval difficulties. Many organisations depend on outdated, fragmented infrastructure and legacy search tools not built for today’s complex, hybrid workloads necessary for AI-driven applications. This often leads to slow, inefficient, and sometimes inaccurate search results, hindering developers and AI agents from accessing vital information. TopK seeks to transform this experience.
“AI agents are becoming increasingly relevant in the workplace, but they require accurate knowledge to execute their tasks correctly and reliably. Additionally, most useful information is unstructured, so traditional information retrieval systems fall short for this use case. TopK’s platform addresses this problem by integrating vector-based retrieval for unstructured and multi-modal data with keyword and metadata filtering to provide highly relevant context, thus improving the overall reliability of agentic systems,” explains Marek Galovič, CEO & co-founder of TopK, in an exclusive conversation with TFN.
Today, TopK has secured $5.5 million in seed funding from Earlybird, KAYA, Irregular Expressions, and several notable angel investors. This funding will enable TopK to grow its team and speed up the development of its enterprise solutions, which are aimed at supporting web-scale search applications across various industries. When asked about the company’s valuation, Galovič declined to provide details but confirmed, “It’s $5.5 million” as the total funding raised so far.
Founders on a mission to reimagine search
TopK’s founding team combines strong technical skills with a shared vision for the future of search. Marek Galovič and Jerguš Lejko, who have known each other since high school, reunited after years of developing large-scale systems and AI tools to fundamentally rethink search. Their motivation was driven by their personal growth and a clear market demand. “It was both – the personal developer experience of the founders and the identified market need for a more flexible solution,” explains Galovič.
The founding team is deliberately diverse, comprising four members of varying genders, backgrounds, and sexual orientations. Their goal is to create a unified platform for relevant, intelligent search across all enterprise data, removing the reliance on patchwork solutions and allowing organisations to harness AI’s full potential.
The founders aim high: to establish the ultimate query layer that drives search for the world’s most advanced applications, facilitating both human users and AI agents in discovering and comprehending information with unmatched speed and precision.
Behind TopK: A unified, AI-native search engine
TopK’s core technology features a “true hybrid retrieval” engine that integrates vector search, keyword matching, and filtering into a single, cloud-native platform. This consolidated method provides development teams extensive control over retrieval and ranking, enabling them to customise results for their particular domain, such as e-commerce, finance, healthcare, or law.
“Existing solutions on the market offer hybrid retrieval by querying multiple indexes and then merging the results, for example, using rank fusion. This degrades the quality of results and is inefficient, since data has to be indexed and queried multiple times. We built our solution from the ground up, including storage format and query engine, which allows us to combine various types of retrieval in a single query and on top of the same data. This way, we get much higher quality results, up to 5x faster and 10x cheaper,” Galovič explains.
Unlike competitors like Elasticsearch, Turbopuffer, or LanceDB, TopK’s integrated system allows complex queries to be run in a single search request, greatly enhancing accuracy and efficiency. The platform is built for high performance and cost-effectiveness, making it suitable for large-scale production environments where speed and quality are crucial.
Building for the future: talent and expansion
Looking ahead, TopK focuses on enhancing its enterprise-grade features and broadening its capabilities to provide intelligent search solutions. The company is expanding its team in Prague, tapping into Europe’s excellent talent pool in databases, distributed systems, and cloud infrastructure.
TopK seeks domain experts and technical leaders who are eager to take ownership of key verticals and foster innovation across the platform. The founders are also open to designing partnerships with organisations tackling complex search challenges. They aim to push the limits of enterprise search and ensure the platform evolves to meet demanding real-world needs.
With strong support from top investors and a clear vision, TopK is ready to change how enterprises search and interpret their data, empowering developers and AI agents in the fast-moving AI-native applications landscape.
“Our goal is to provide a vertical search solution for AI agents, primarily focused on enterprise use cases. To this end, we plan to add enterprise features, integrations with existing data systems, and make TopK’s retrieval engine support web-scale workloads with billions of documents,” concludes Galovič.
