A year ago, using SurrealMCP meant installing and running your own server. With the new managed SurrealDB MCP, simply use one URL, sign in, and your assistant can deploy instances, query data, and recall memory across chats. Get started. 👉 sdb.li/3TuT8PN

Did you know Spectron lets separate teams share one Context without seeing each other’s memory? Scope tags on remember() and recall() partition data by team, org, project, and so on. Learn more in our docs. 👉 sdb.li/4fOO6EZ

Semantic search usually means two hops: your app embeds the text, then queries the database. @martinschaer.bsky.social shows how one SurrealQL function does both, plus a graph traversal, in one query. No vectorisation service needed. Take a look. 👉 sdb.li/4w5tgYb

Need tools, libraries, and integrations built around our multi-model database? SurrealDB Labs brings together official and community-built resources to help you build faster. Dive in. 👉 sdb.li/4vYwpc8

What if you could watch a graph as it grows and fire off an action the instant a meaningful pattern completes? Community member Cyril Scetbon explores exactly that: concept, configuration, implementation, and where things could go from there. 👉 sdb.li/4x8eVuV

How do you keep your agent’s knowledge layer fresh, without full rebuilds? On 30 July, Tobie Morgan Hitchcock, @badmonster0.bsky.social (@cocoindex.bsky.social ex.bsky.social), and @martinschaer.bsky.social (SurrealDB) show you how. Check out the full details. 👇

Stream #37: Building a living knowledge layer for your agents

YouTube video by SurrealDB

youtube.com

Almost every SaaS app needs the same plumbing: orgs, roles, permissions you write once, get subtly wrong, and patch for years. SurrealKit ships Organizations and Teams as tested SurrealQL you opt into. Learn more from Chiru Boggavarapu. 👉 sdb.li/4fwT5v3

Struggling with a fragmented backend architecture? Learn why Aspire Competitions migrated to our multi-model database, and how it helped cut complexity and speed up development. 👉 sdb.li/3R8BxMw

Did you know Spectron lets you store context with one line of code? The remember_many() method extracts facts automatically and assigns them to the right user. Learn more in our docs. 👉 sdb.li/4fgQSTe

Ready to get hands-on with SurrealDB? This tutorial gives you the practical skills to build a movie database while mastering our multi-model database and SurrealQL. Start building today. 👉 sdb.li/4kpSpbK

RAG can retrieve information, but it can't determine authority, precedence, or approvals. Tobie Morgan Hitchcock explains in Verdict why context, not bigger prompts, is what makes agentic AI reliable. Read the full piece. 👉 sdb.li/4eW3QqO

"Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo" is technically valid English, but it's confusing. Our own @mithridates.bsky.social uses SurrealQL and Surrealist's graph visualisation to show exactly how it works. 👉 sdb.li/4vU0N8A

Every new AI capability shouldn't mean a new service to manage. Cobrainer avoided fragmenting its stack by running graph, vector, full-text, and memory for its AI agent on one engine, our multi-model database. Read the full story. 👉 sdb.li/3R2eeE0

At Camp AI this week, hosted by @auth0byokta.bsky.social, @martinschaer.bsky.social will showcase the benefits of having a multi-model DB for your agents: rich context through a single text-to-surql tool call, and only one query to build the context. Full event details. 👉 luma.com/camp-ai-july...

Did you know our multi-model database lets you define record user access and per-table permissions directly in SurrealQL? Authentication lives inside the database, not an external service. Learn more in our docs. 👉 sdb.li/4wppgSR

Context layer, semantic layer, knowledge graph: three terms AI teams can confuse, often building them as three disconnected systems. SurrealDB handles all three in a single query, in a single round trip. Learn how. 👉 sdb.li/3SWNN3e

Developer Office Hours is back today. Join us at 4 PM BST on Discord, where host Chiru Boggavarapu will walk you through two SurrealKit features: rollouts (shipped) and the upcoming multi-database feature. See you there. 👉 sdb.li/4ffMfKi

Production traffic, uptime commitments, AI agents that can't stall. Scale, the new tier of SurrealDB Cloud, our fully managed database service, is for teams that need high availability and scale in production. Learn more. 👉 sdb.li/4ghyYSz

Why let Replit Agent improvise a backend when you already own one? Chiru Boggavarapu walks through connecting Replit Agent to SurrealDB via MCP, ending with a real interface running on your own data. Try it for yourself. 👉 sdb.li/4p2euQ7

Did you know SurrealDB gives you ACID transactions across document and graph writes? Every write commits or rolls back together with no partial state, including across multi-statement manual transactions. Learn more in our docs. 👉 sdb.li/4ycbQvo

Meet SurrealQL, our powerful SQL-like query language built for our multi-model database. Intuitive for developers, it makes querying across data models straightforward. Start your journey of learning SurrealQL today. 👉 sdb.li/3Khkft6

💜 Dev quote of the week.

Schemic, built by community member Manuel Sánchez, lets developers define schemas in the Zod API and generate native DDL, types, and migrations. It ships with SurrealDB driver support; check out the GitHub repo. 👉 github.com/schemichq/sc...

Most agentic coding tools improvise the backend. SurrealDB changes that. Chiru Boggavarapu shows how connecting Lovable to SurrealDB via MCP makes your database the system of record and source of truth the AI builds against. See how it works. 👉 sdb.li/44qOhRG

Surrealist 3.9, our latest major UI release, introduces a complete design overhaul, a new datasets browser and data manager, an improved record inspector, and more. Explore the highlights. 👉 sdb.li/3Qk4iWl

Did you know SurrealDB lets you query graph, vector, and temporal data in a single round trip? Learn more in our docs on how our multi-model database lets you put an end to ORMs, joins, and glue code between datastores. 👉 sdb.li/4xyo5BY

RAG works well for retrieval, but agentic AI needs more. The missing layer is usually a world model, and knowledge graphs are one way to provide it. Tobie Morgan Hitchcock explains why in Techstrong AI. Read the full piece. 👉 sdb.li/4gukpLl

Developer Office Hours returns today at 4 PM BST on Discord. Host Chiru Boggavarapu welcomes community members Manuel Sánchez and Drew Ridley, showcasing projects SurrealDB Zod and SurrealGuard. See you there. 👉 sdb.li/4gvyrfP

Building AI workloads on Nebius? SurrealDB is on the Nebius AI Cloud Marketplace. Document, graph, and relational capabilities in one unified data layer, designed for agent memory and agentic workflows. Find out more. 👉 sdb.li/4fOElbG

Attending @aidotengineer.bsky.social World’s Fair in San Francisco and looking to go deeper on agents? Join us on June 30 when Tobie Morgan Hitchcock and engineers from @vercel.com and awscloud.bsky.social tackle building agents that hold up in production. Register. 👉 sdb.li/4uJ3Te7