Startup Keewano launches agent-focused database and pulls in $12M in funding
Hetz VenturesCompany
Andreessen HorowitzInvestor
Keewano, the Tel Aviv‑based AI‑agent database startup, closed a $12 million venture round on Sept. 15, 2026, led by Hetz Ventures and Andreessen Horowitz’s a16z Speedrun, to fund the launch of KeewanoDB.
Keewano, a Tel Aviv‑based startup (officially Sandstorm Ltd.), raised $12 million in a venture round led by Hetz Ventures and Andreessen Horowitz’s a16z Speedrun on Sept. 15, 2026. The capital will back the commercial rollout of KeewanoDB, the company’s event‑oriented database built for artificial‑intelligence agents.
Deal Terms
The undisclosed‑type round totals $12 million, with Hetz Ventures and a16z Speedrun as the lead investors and participation from additional, unnamed backers. The funding is earmarked for product engineering, go‑to‑market expansion, and hiring across engineering and sales. No valuation or ARR figures were disclosed.
Product Overview
KeewanoDB stores events—such as customer actions, device signals, or transaction records—in the order they occur, grouping them around an entity. The design eliminates the need for traditional ETL pipelines; data remains live on solid‑state drives and is queried directly by AI agents. The system runs in Docker containers orchestrated by Kubernetes, leverages vectorized CPU instructions, and claims the ability to scan 250 million events in under 0.5 seconds, though independent verification is pending.
The platform supports Parquet and Apache Iceberg formats and can ingest streams from Apache Kafka. Pricing is based on the number of active entities rather than per‑event counts, a model that customers reportedly translates into token‑usage savings of roughly 84 % versus querying Snowflake or BigQuery directly. Keewano also bundles a Lua‑based in‑database acceleration engine and a semantic layer that auto‑generates a business view of incoming events.
Keewano plans to ship two offerings: the core KeewanoDB for agentic analytics and a higher‑level analytics layer that includes a Signal agent for hypothesis testing and real‑time operational decisions. The funding will accelerate these launches, expand the engineering team, and fund sales outreach to enterprises building AI‑driven products.
Why It Matters
The infusion of $12 million gives Keewano the runway to position KeewanoDB as a niche alternative to traditional data warehouses for AI‑centric workloads. By pricing on active entities and delivering on‑database reasoning, Keewano can undercut the compute‑heavy querying costs that Snowflake, Databricks, and Google BigQuery charge for similar analytics, potentially prompting those vendors to revisit pricing or add native agent support.
For incumbents that rely on relational or columnar stores, Keewano’s event‑first architecture forces a rethink of data modeling for AI agents. If Keewano can validate its performance claims at scale, developers may favor it for real‑time decision loops, accelerating the shift toward agent‑driven product features and eroding the market share of legacy warehouses in that segment.
Key Points
- Keewano closed a $12 million venture round on Sept. 15, 2026
- Lead investors are Hetz Ventures and Andreessen Horowitz’s a16z Speedrun
- KeewanoDB is an event‑oriented database designed for AI agents, eliminating ETL pipelines
- Pricing is based on active entities, with reported token‑usage savings of ~84 % versus Snowflake/BigQuery
- The round will fund product launch, engineering expansion, and go‑to‑market efforts
Analysis
The $12 million raise places Keewano in the early‑stage SaaS valuation band where investors typically apply 10‑15 times forward‑looking ARR for AI‑enabled data platforms. Assuming the company can achieve $2‑3 million ARR within the next 12‑18 months, the implied multiple would be in line with recent AI‑infrastructure deals, suggesting a modest premium for its agent‑centric approach. The funding underscores a broader trend: investors are betting on specialized databases that serve AI agents rather than human analysts, a niche that could command higher gross margins as compute shifts from cloud‑wide warehouses to edge‑proximate, low‑latency stores. For operators, the capital enables rapid scaling of engineering resources to meet performance benchmarks, while the pricing model based on active entities offers a predictable revenue stream that can boost net‑revenue retention as customers expand agent usage. Investors will watch Keewano’s ability to convert early adopters into recurring revenue and to demonstrate measurable cost‑savings for AI workloads, metrics that will drive future valuation rounds.