Unifying Real Time Search and Analytics in SQL

Modern software architectures often suffer from fragmenting operational data across separate engines for text search, vector retrieval, and analytical queries. Engineers frequently spend significant resources setting up complex data pipelines to push records from transactional databases into specialised search clusters and analytical data warehouses. This separation creates synchronisation latency, increases cloud infrastructure bills, and introduces frequent point-of-failure risks into production environments. Berlin-based software venture SereneDB GmbH was established to eliminate this structural duplication by creating a unified data engine capable of handling multiple workloads concurrently.

The company was founded in Berlin in 2025 by Alexander Malandin, Andrey Abramov, and Valery Mironov to tackle these persistent data pipeline bottlenecks. Drawing on extensive technical experience in search analytics and distributed system architecture, the founding team designed an open-source distributed database from the ground up. Their unified engine combines full-text BM25 keyword search, vector similarity matching, and columnar analytical SQL processing within a single execution layer. By enabling technical teams to execute hybrid queries alongside analytical aggregations without duplicating underlying storage, the system significantly simplifies core enterprise infrastructure.

At the heart of the platform is a distributed database engine designed to ingest, update, and query structured and unstructured data in milliseconds. Rather than forcing developers to adopt proprietary query syntax, SereneDB provides a PostgreSQL-compatible interface that integrates directly with existing SQL tools, business intelligence dashboards, and application frameworks. Software engineers can compose standard SQL queries that execute BM25 keyword matching, vector nearest-neighbour searches, and analytical GROUP BY operations in a single query execution pass. The engine also features zero-ETL functionality for querying remote databases and object storage files directly where they reside.

The practical use cases span multiple sector requirements, including security operations, financial fraud monitoring, digital commerce, and artificial intelligence retrieval workflows. Security teams use the platform for high-throughput log analysis and real-time threat hunting, while commercial applications execute fast product catalogue searches with typo handling and facet filtering. Furthermore, artificial intelligence engineering teams employ the database for retrieval-augmented generation workloads, supplying large language models with accurate context using scoped SQL permission boundaries. In June 2025, the venture secured 2.1 million euros in seed funding led by Entourage, with participation from High-Tech Gründerfonds.

Looking ahead, the team is expanding its open-source ecosystem, conducting public benchmarks against traditional search and OLAP engines, and preparing to present its platform at Web Summit 2026. Supported by a company valuation of 8.807 million euros following its pre-seed funding round, SereneDB continues to enhance its distributed query execution mechanisms and developer tools, including the SereneUI interface. By bringing full-text search, vector embeddings, and analytical SQL into a unified engine, the Berlin startup aims to help engineering organisations streamline data management for real-time applications.

Find out more at serenedb.com