The serverless cloud data warehouse built on DuckDB, for humans, agents and applications
HQ: Seattle, Washington, USA (Amsterdam engineering site; six AWS regions, Azure waitlist opened October 2026) | Founded: 2022 | Employees: ~160 (est.: LinkedIn company page ~164; PitchBook 159; Tracxn 160) | Stage: Series B (Private) — ~$133M raised; last round May 2025 | Website: https://motherduck.com
MotherDuck is a Seattle-based cloud data warehouse company built on DuckDB, the open-source analytics database created by Hannes Mühleisen and Mark Raasveldt at CWI in Amsterdam. It was founded in 2022 by Jordan Tigani, a founding engineer on Google's BigQuery and later chief product officer at SingleStore, who decided DuckDB needed a serverless cloud sibling. DuckDB Labs, the database's commercial arm, became a co-founder and shareholder. The team met in Seattle in June 2022 and left stealth that November with $47.5M at a $175M post-money valuation.
Architecturally MotherDuck is not a distributed warehouse. It runs DuckDB as a 'hypertenancy' service: every user, application or agent gets an isolated serverless instance — a Duckling — starting in about 100 milliseconds and stopping when idle, with no shared clusters or noisy neighbours. Its dual-execution planner decides per query whether data runs locally or in the cloud, so an analyst can join a laptop CSV to a cloud table in one statement. Execution is single-node: no shuffling overhead, and a deliberate focus on workloads under ~10-20TB.
The platform went to preview in June 2023, dropped its waitlist alongside a $52.5M Series B led by Felicis at a $400M valuation that September, and reached GA in June 2024. It sells a free Lite plan, a $250-per-org Business plan and custom Enterprise contracts on per-second metering across six AWS regions. Customers include Together AI, UDisc and FinQore; dbt, Hex and Tableau integrate natively.
Since mid-2026 it has repositioned for the agent era — MCP server, Flights pipelines, Dives visualizations and AI functions billed in AI Units — while acquiring Tower Computing (Aug 2026), launching commercial DuckDB support (Sep) and opening an Azure waitlist (Oct).
MotherDuck is the cleanest bet in the market that the cloud data warehouse needs simplification more than it needs scale. Its premise, set out by CEO and BigQuery founding engineer Jordan Tigani when the company surfaced in 2022, was that 'big data is dead': most businesses have tens of terabytes rather than petabytes, and one fast node beats a distributed cluster. MotherDuck runs DuckDB — the MIT-licensed in-process analytics database created at CWI in Amsterdam — as a serverless service where every user, customer and agent gets an isolated 'Duckling' that starts in about 100ms and stops when idle, while the same query planner splits work between the laptop and the cloud.
The funding has been steady but modest by warehouse standards: $47.5M at a $175M post-money valuation in November 2022 (Redpoint seed, a16z Series A), $52.5M led by Felicis at $400M in September 2023, and a $33M extension in May 2025 that brought total capital to roughly $133M — less than ClickHouse raised in a single round. No revenue figure has ever been published; third-party estimates range from $5.8M (2023) to about $12.6M (est.).
The ground shifted in August 2026 when DuckLabs, the company employing DuckDB's core engineers, joined AWS. The code stays MIT-licensed under the non-profit DuckDB Foundation, and MotherDuck publicly backed the deal, but its engine's authors now work for a hyperscaler selling competing warehouses. MotherDuck's answer is product rather than ownership: an agent-native layer (MCP server, Flights, Dives, AI functions), a commercial open-source DuckDB support business launched two weeks after the AWS announcement, an Azure waitlist, and the Tower acquisition to run agent-built pipelines. The upside is that agents already generate 29x more queries than humans on its platform and need cheap isolated compute; the ceiling is single-node architecture below ~10-20TB, free DuckDB as a substitute, and rivals with several times its capital.
Browse: Home | Blog | About | All Startups A-Z | View full profile on StartupWiki