Keewano has announced the general availability of KeewanoDB, a database purpose-built for machine reasoning at scale. The Tel Aviv based company also disclosed total seed funding of $12 million led by Hetz Ventures, with participation from a16z speedrun, Remagine Ventures, DIG Ventures, and angel investors. The platform is designed to provide AI agents with the complete, ordered, and contextual information that traditional analytics databases were never built to handle.
Addressing a Gap in Analytics Infrastructure
Conventional analytics infrastructure has long been optimized for human users rather than machines, storing events as flat rows, columns, and precomputed aggregates. This structure forces teams to reconstruct event order and context at query time, making complex behavioral questions slower and more expensive. Many organizations therefore capture only a few hundred distinct event types and discard the surrounding detail that AI agents need for grounded reasoning.
How KeewanoDB Works
KeewanoDB takes a fundamentally different architectural approach by keeping each entity's complete event sequence together and in order. The system stores every action a user, device, or AI agent undertakes so that it remains live, ordered, and contextually available. In practice, the company reports that a quarter of a billion events can be queried in under half a second, returning context-ready results that agents can reason over immediately.
Use Cases and Pricing Model
With KeewanoDB, AI agents can investigate questions that dashboards cannot easily answer, including why a customer churned and what sequence of behaviors led to an outcome. The system also helps agents identify shared patterns among users, distinguish one group from another, and proactively surface emerging risks. Because KeewanoDB does not use per-event pricing, organizations can capture more data and run more queries without predictable cost escalation.
Deployment and Integration
Keewano Cloud is a fully managed service that can connect alongside an existing data warehouse or replace it entirely through standard integrations. Organizations can plug in their own AI agents or use Keewano's built-in agent capabilities, giving teams flexibility in how they adopt the system. This design is intended to reduce storage and processing overhead while supporting agent-scale throughput across large event histories.
Founding Team and Backing
Keewano was co-founded in 2024 by Mark Kardashov, Dima Karger, Pavel Bibergal, and Vitaly Bukhovsky. The leadership team brings deep experience in analytics, B2B software, and gaming, with Kardashov and Bukhovsky having built and sold two companies together. Their previous roles include senior positions at Plarium and leading game studios, providing direct insight into high-volume event data at scale.
Investor Perspective
Investors are positioning KeewanoDB as a response to the infrastructure gap exposed by the current wave of AI adoption. Hetz Ventures led the seed round and sees machine reasoning as a workload that conventional analytics tools were not designed to support. The participation of a16z speedrun, Remagine Ventures, DIG Ventures, and angel investors further signals confidence in the company's architectural approach.
Vision for the Future
Keewano's founders view the current database landscape as a fork rather than an incremental improvement. They argue that AI agents require a fundamentally different data architecture, just as high-speed rail required new infrastructure rather than faster trains on old track. This vision positions KeewanoDB as a new line built specifically for how machines reason, rather than a patch on human-centric systems.
KeewanoDB enters the market as a new option for organizations that want to support AI agents with complete, ordered, and context-rich event data. Its general availability and $12 million seed round mark a meaningful step toward replacing analytics systems designed for human dashboards. As AI reasoning continues to evolve, purpose-built databases like KeewanoDB could become an important layer in the modern data stack.