Senior Software Engineer @SurrealDB
Harare, Zimbabwe
Joined October 2013
- Tweets200
- Following185
- Followers186
- Likes682
Rushmore Mushambi retweeted
Building agent memory? Start with the right ingredients: every fact with its source, permissions that stick, and corrections that keep history. The SurrealDB Agent Memory Cookbook offers practical end-to-end guides.
Pick a recipe. 👉 sdb.li/4BgWxmx
Rushmore Mushambi retweeted
Just a reminder, there's another release from @SurrealDB this week.
Rushmore Mushambi retweeted
In the real world, you don't get to tell your agent what kind of question it's about to be asked. A benchmark shouldn't either.
So ours doesn't. 77.2% on BEAM at 1M, 57.1% strict, every number with its judge.
Rushmore Mushambi retweeted
Memory or accuracy? Yes.
SurrealDB Agent Memory: memory for AI agents, every fact with its source.
77.2% on BEAM, the 1M-token memory benchmark. 57.1% strict, every rubric item met in full. No other vendor we checked publishes that.
Learn more: sdb.li/4zqTxCi
Stop choosing.
Rushmore Mushambi retweeted
Speed or durability? Yes.
We're replacing RocksDB.
New persistent storage engine: reads 8.0x and updates 6.1x faster than RocksDB. Reads 73.7x faster than SlateDB (on disk, not object storage).
Run it yourself.
Stop choosing.
sdb.li/3VE7i24
Rushmore Mushambi retweeted
We're live on Product Hunt with Spectron, the memory and knowledge layer for AI agents built on SurrealDB. Come share your thoughts and drop your questions in the thread; we'll be there all day. 👉 sdb.li/4fnDYEK
Rushmore Mushambi retweeted
1/ A few years ago we made a bet that the database should be multi-model in one engine, in one transaction. Most people thought that was a niche idea. Agent memory is where that bet pays off.
Rushmore Mushambi retweeted
Agents are getting good at acting. They're still bad at remembering. Spectron, the memory and knowledge layer for AI agents built on SurrealDB, addresses this. Early access opens today. Learn more. 👉 sdb.li/4fXn8g4
Rushmore Mushambi retweeted
We have been working on something. Building AI agents? Tomorrow is for you.
3 June 2026.
Rushmore Mushambi retweeted
1/ Databases are entering a weird new era.
Not because of AI hype.
Because for the first time in decades, databases are being asked to behave less like storage engines… and more like memory systems.
Today we shipped SurrealDB 3.1.
And buried inside the release is something important: DiskANN.
Rushmore Mushambi retweeted
SurrealDB 3.1 is here and builds on the foundations of 3.0, with a focus on stability, a second approximate-nearest-neighbour vector index in DiskANN, GraphQL upgrades, and significant security hardening.
Explore the highlights.👉 sdb.li/3Q7X4Ey
Rushmore Mushambi retweeted
#NVIDIAGTC is live, and we're at booth #4026! Come meet founders @tobiemh & @jaimemh, and chat with Matthew Penaroza & Ignacio Paz to learn why SurrealDB is becoming the go-to database for AI agents, handling core state, contextual memory, and connected data in one place. 👋
Rushmore Mushambi retweeted
We rebuilt SurrealDB’s core engine in our latest major release, unlocking significant performance gains over prior versions. Benchmarks show graph queries 8–22× faster, with more improvements already on the roadmap.
Dive into the details. 👉 sdb.li/4kOBVuc
Rushmore Mushambi retweeted
SurrealDB 3.0 launches with $23M in new funding and a pitch to replace multi-database RAG stacks with a single engine that handles vectors, graphs, and agent memory transactionally.
venturebeat.com/ai/surrealdb…
Rushmore Mushambi retweeted
Today we are releasing @SurrealDB 3.0. With this version, we have strengthened the core architecture to better support modern, AI-driven applications. Rather than treating graph, document, relational, and vector capabilities as separate concerns, 3.0 brings them together within a single engine and unified query layer. The goal is straightforward: reduce fragmentation and provide developers with a consistent foundation to build on.
AI systems place different demands on infrastructure. They require structured data, semantic search, and rich relationship traversal to operate reliably at scale. SurrealDB 3.0 supports these patterns natively, allowing context and data to exist within one coherent system rather than being distributed across multiple services.
Significant work has gone into improving stability, performance, and production readiness. We have refined how data is stored and processed, strengthened transactional guarantees, and simplified the overall developer experience. The result is a database that is more predictable under load and easier to integrate into real-world deployments.
This release would not exist without the dedication of the team and the thoughtful feedback from our community. We are grateful for the role both have played in shaping 3.0.
If you are building intelligent, data-intensive systems, SurrealDB 3.0 is ready to explore. You can watch the launch and start building today.
surrealdb.com/3.0
Rushmore Mushambi retweeted
SurrealDB 3.0 is coming, mark your calendar! On February 17, 2 PM GMT we’ll unveil our latest major release, sharing what’s changed under the hood and how it reshapes the way you model and query data, especially for modern, AI-driven applications. 👉 youtu.be/USg8ZQC5mQc
Rushmore Mushambi retweeted
A schema defines your database’s structure and rules. Early definition ensures consistent data storage, linking, and behaviour throughout your project. Some tips from our own @mithridates show you how to design one that works for you. 👉 sdb.li/4ly1PkX
Rushmore Mushambi retweeted
Our Rust SDK is the primary method for interacting with our multi-model database from client–side, server-side apps, systems, APIs, embedded systems, and IoT devices. Check out 7 tips and tricks for optimising your use of the SDK.👇 sdb.li/4ljfYmI
Rushmore Mushambi retweeted
Eager to learn how our multi-model database stacks up against other databases, whether it be SQL, NoSQL, and others? Join us on this benchmarking journey. 👉 sdb.li/4hs28vY
