Startup of the Day: Czech Recombee, an AI platform for content personalization

Recombee is a Czech AI startup developing a platform for content personalization and recommendations. Founded by Pavel Kordik, Tomas Rehorek, Gabriela Takacova, Ondrej Fiedler, Tomas Barton, Antonin Kral, the company works with media, streaming services, and marketplaces, providing real-time recommendation and search technologies for clients in multiple countries.

In the Startup of the Day column, Recombee’s team shares more about the company’s technology, current traction, and upcoming developments.

The Startup of the Day column on AIN.Capital is dedicated to tech projects from all sectors that originate from the CEE countries. If you would like to introduce your project, please fill in the questionnaire.

Tell us about your startup. How does it work?

Recombee is an AI-powered personalization platform that enables digital businesses of all kinds to deliver highly relevant content and product recommendations to their users. We provide a sophisticated recommendation engine as a service, the kind that only used to be accessible to major corporations like Amazon or Netflix. 

The platform uses over 100 built-in proprietary machine learning models to analyze and connect the dots between user behavior, item similarities, and business goals, all of which are then used to generate real-time recommendations. 

Outside of recommendations, Recombee’s personalized search improves the user experience by making content easier to discover and navigate, while its advanced analytics give companies actionable insights into what resonates with audiences, supporting smarter editorial and product decisions. 

User behavior like clicks, views, and purchases are analyzed in real time and enhanced with collaborative filtering, content-based models, reinforcement learning, and proprietary architectures. The end product is truly relevant results that match each user’s interests and context across different content types and use cases.

How did you come up with the startup’s idea? What was the reason/motivation behind it?

We’re all machine learning researchers who started working together at Czech Technical University. We essentially realized that the recommendation technologies used internally by giants like Amazon and YouTube could be made accessible to businesses of all sizes. Our idea was to democratize advanced AI and make world-class recommendation technology available as a SaaS product.

How long did it take to reach the prototype or MVP? What did you encounter?

Early versions of our product were highly technical and required API calls without a user interface. Our MVP delivered basic collaborative filtering recommendations built from collected user-item interactions, that are still foundational to our product. 

One of the first real challenges we faced was making the solution accessible to non-developers by creating an admin UI. We also spent a lot of time and energy convincing businesses of the value of personalization at a time when AI and recommender systems were poorly understood.

When exactly did you launch your product? Or when the launch is planned?

We founded the company in 2015. Early versions of the product were already in use that year with a few clients, before Gabriela Takáčová joined as co-founder to build the business operations and drive growth, taking Recombee from a handful of early adopters to usage on thousands of sites worldwide, including some of the world’s biggest platforms.

Tell us about the stratup’s business model. How do you monetize your product?

Recombee uses a subscription-based, pay-as-you-go model. Pricing is based on usage (number of recommendations, catalog size, active users) with multiple tiers (Free, Standard, Plus, Pro, Premium). Higher tiers include advanced support and custom model adjustments.

What are your target markets and consumers?

Recombee serves mid-market and enterprise companies with large catalogs and user bases. Key industries are news & media, video and audio streaming, deal aggregators, real estate, gaming, and e-commerce/marketplaces. Current focus is especially strong on media and publishing. We’ve made real inroads there thanks to clients like The Telegraph, DAZN, Fox News, 9gag, and Pepper.

If the startup has already launched the product, what are the results: metrics, income, or any clear indicators that can be evaluated.

• We handle over 30,000 recommendations per second and recommends over a billion items daily

• Clients have seen up to +50% CTR and +37% post views

• Case studies: 9gag increased user session duration and return rates; DAZN and The Telegraph use Recombee for large-scale, real-time personalization.

• The company generates approximately $6M in annual recurring revenue

What about your team? How many people are working in the startup? If you’re looking for new employees, indicate whom exactly.

Recombee employs over 50 people. We are a group of hi-tech enthusiasts, and we constantly look for open-minded, creative and passionate people to join our team in the vibrant city of Prague.

Have you already raised any investments? Provide us with more details on each funding round: the amount, investors, the purpose of the investment.

We are fully independent and bootstrapped.

What’s next? Tell us about your future plans.

We’re currently preparing to launch in-UI A/B testing to help customers easily optimize their recommendations. We’re expanding LLM-powered features such as homepage row creation and emerging topic detection to streamline curation and surface trends.

Strategically, we’ve shifted focus somewhat toward the global media and publishing markets, all while continuing to grow in video, marketplaces, and streaming. 

Our long term mission is, however, basically unchanged. We’re here to empower businesses with world-class personalization tools that combine AI automation with human editorial control.