Skip to main content
AI Market Watch
Loading...

Keewano

Category: AI Infrastructure

Keewano (Sandstorm Ltd.) builds KeewanoDB, an event-sequence database purpose-built for AI agents and machine reasoning at scale. Keewano was founded in 2024. The company is led by Mark Kardashov. Based in Tel Aviv, Israel. Team size: 11-50. Total funding raised: $12M. Latest round: Seed. Key investors include Hetz Ventures, a16z Speedrun, Remagine Ventures, DIG Ventures.

Founded
2024
Headquarters
Tel Aviv, Israel
Team size
11-50
Total funding
$12M

Value proposition

KeewanoDB stores each entity's complete event sequence together and in order, giving AI agents full context to reason over data — answering open-ended questions traditional databases (built for humans asking predefined queries) were never designed to handle.

Products and solutions

KeewanoDB (Local & Cloud) — event-sequence database for machine reasoning, Keewano Analytics — AI analytics agent layer, Keewano Ask for Slack, Keewano MCP for Agents, Keewano Signal — autonomous threat/opportunity detection agent.

Unique value

First database built from the ground up for machine reasoning: keeps every event intact, in order, in context, live the moment it lands — no ETL, no pipelines, no data lakes. Claims ~250M events queried in under half a second and up to 84% LLM token cost reduction via in-database acceleration.

Target customer

Engineering, product, analytics, operations, and AI teams running high-volume event workloads and production AI systems; initially gaming analytics, expanding to enterprise verticals.

Industries served

Gaming (origin), enterprise SaaS, product analytics, fraud/churn detection, real-time operational analytics

Technology advantage

Event-oriented storage format (~4 bytes/event on SSDs) with contiguous ordered sequences; distributed multi-sharded design on Kubernetes/Docker; SIMD/GPU acceleration; in-database Lua acceleration engine via Model Context Protocol (MCP); semantic layer auto-constructs business context; supports Parquet, Apache Iceberg, Kafka, Snowflake, BigQuery ingestion.

How they differentiate

Unlike row/column analytics databases optimized for predefined human queries, KeewanoDB is architected around complete ordered event sequences so AI agents can investigate unanticipated questions without data reconstruction or ETL pipelines, with reasoning executed in-database.

Main competitors

MotherDuck/DuckDB, Tinybird, Rockset, traditional analytics databases (Snowflake, BigQuery)

Key partnerships

a16z Speedrun (Cohort 003), Hetz Ventures, Remagine Ventures, DIG Ventures

Major milestones

Founded 2024 by ex-Plarium/TestProject/Devalore founders, out of stealth Sept 15, 2026 with $12M Seed led by Hetz Ventures and launch of KeewanoDB, a16z Speedrun Cohort 003.

Market positioning

Early-stage (out of stealth Sept 2026) entrant in the emerging "database for AI agents / machine reasoning" category, backed by top-tier VCs (Hetz Ventures, a16z Speedrun).

Geographic focus

Global (headquartered in Tel Aviv, Israel)

About Mark Kardashov

Ex-General Manager at Plarium (20+ years in gaming/business ops); co-founded TestProject (acquired by Tricentis, 2019) and Devalore (acquired by Abra, 2022).

Latest news about Keewano

More AI Infrastructure companies

Official website: