TigerBeetle

Market position

TigerBeetle positions itself as a complement to, rather than a replacement for, general-purpose (OLGP) and analytical (OLAP) SQL databases, arguing that OLTP — writing fixed-size integers such as orders and payments — is a distinct workload from reading variable-length strings. It explicitly benchmarks itself against Postgres and names Postgres, MySQL and SQLite as the incumbent general-purpose designs it considers unsuited to modern transaction volumes. Public traction signals include 16.9k GitHub stars and 887 forks.

Key differentiators claimed are a purpose-built, fixed debit/credit schema with strict serializability that moves OLTP invariants such as idempotency, isolation and two-phase commit out of application code; batching of thousands of transactions into a single query to avoid lock contention; deterministic low tail latency through static allocation and Direct I/O/io_uring; a replicated six-machine architecture that scales up rather than out; and full open-source availability under Apache 2.0 without contributor licence agreements.

Technology

TigerBeetle co-designs its global consensus protocol and local storage engine as a single system optimized for transaction throughput per CPU cycle under high write contention. Design elements cited include batching of up to 8,189 transactions per query, elimination of row locks, static memory allocation, zero copy, zero deserialization, Direct I/O and io_uring for predictable P100 tail latency, no garbage collection or out-of-memory crashes, and a tiered storage engine spanning L1/L2/L3 CPU caches, RAM and NVMe. Clusters scale up on six replicated (never partitioned) machines. The codebase is written with Zig tooling in the repository and follows a documented engineering methodology, TIGER_STYLE.md. A security contact (security@tigerbeetle.com) is published for vulnerability disclosure.

Go-to-market

Distribution is developer-led: a downloadable single binary via curl one-liners for Linux, macOS and Windows, extensive public documentation at docs.tigerbeetle.com, a GitHub repository, and client libraries for major languages. Community channels include a monthly newsletter, a YouTube channel (including talks from the Systems Distributed conference), Twitch streams, X, Bluesky and GitHub. Technical talks (e.g. "Redesigning OLTP for a New Order of Magnitude" at QCon SF, and a video introduction with The Primeagen) are used for awareness. The company stated that Series A funding would be used to grow the team and acquire new customers.

Organizations with transaction-heavy, mission-critical workloads that can be expressed as double-entry bookkeeping — financial services and payments providers, plus sectors the company cites as rapidly becoming more transactional: cloud, energy, ride-sharing, gaming/in-game economies and real-time payments.

Geography

Sources describe TigerBeetle as San Francisco, California-based. The company references global transaction-volume growth in markets such as India and Brazil as market drivers, and supports multi-cloud deployment, but no offices outside the US are named.

History

Spun out of Coil in 2022. In February 2023 the company, then described as San Francisco-based and led by Joran Dirk Greef, Adrian Hope-Bailie and Donovan Changfoot, announced a $6.4M round led by Amplify Partners and Coil with a broad group of angels and advisors. In July 2024 it announced a $24M Series A led by Spark Capital (general partner Natalie Vais), with Amplify Partners, Coil and angels participating, to grow its then eight-person team and win customers. The product has continued on a weekly release cadence, with version 0.17.9 published 6 July 2026.

Risks & controversies

No controversies, litigation or regulatory issues appear in the available sources. Notable profile risks are informational: performance and scale figures are vendor-published and not independently verified in these sources; no customers, revenue or commercial model are disclosed; and the company reported a very small team (eight people) as of mid-2024 while competing against entrenched general-purpose databases.