Search vectors, traverse the graph, and scan recent history in a single call: one externally consistent result instead of three separate systems.
Memory · Data · Execution
The backend your agents run on.
Jennah unifies agent memory, application data, and durable execution in one globally consistent backend, so agents can recall context, query application data, and run reliably without keeping separate databases in sync.
- Single region
- Multi-region geography
- Global
Hands-on access, direct founder support, and input on the roadmap. See the platform
One backend. Always in sync.
Most agent architectures stitch together a vector database, a graph database, a queue, and an application database. Jennah provides one store for all of it, externally consistent at scale, making four guarantees hold across memory, data, and execution at once.
Commit an agent step's log, embeddings, and graph edges together. Once committed, all parts are immediately readable with zero sync lag.
Replay everything an agent knew (logs, vectors, and graph state) exactly as it existed at any past moment.
Memory, application data, and execution state sit behind a single tenant boundary: one system to secure, one system to audit.
Three layers, one backend
Everything an agent needs to remember, store, and execute on a single globally consistent store. Shipped layer by layer.
Remember
Available now
Every kind of agent memory (execution logs, vector recall, and knowledge graphs) in one store, one atomic write, and one snapshot-consistent read.
- Fused semantic search, graph queries, and recent history in one request
- Atomic multi-type writes with read-your-writes consistency across all memory types
- Time-travel an agent's full memory to any past point in time
- Atomic erasure across all memory types in a single transaction
Coming soon: a bi-temporal graph (historical state at time T), search results piped directly into graph traversal, and multi-hop graph queries in a single call.
Store
Available now
Application tables (users, orders, documents) on the same globally consistent backend as your agent memory.
- Declare relational tables on a global, strongly consistent backend
- Inner and left joins across tables in a single query
- Relational filters and exact vector search in one request
- Transactional commits across multiple rows and tables at once
- Agents and applications served through the same API and isolation model
Coming soon: registering relational tables as graph nodes, allowing application data to join graph traversals in a single query.
Run
Coming soon
Durable, crash-safe agent runs that survive worker crashes, deployments, and multi-day approval waits without losing progress or duplicating actions.
- Journaled execution steps: runs resume exactly where they stopped
- Exactly-once side effects across retries and restarts
- Pause for human approval or timers, then resume cleanly
- Bring your own agent loop, or run on ours
Why it matters: durable workflow orchestration and state management handled directly on the same backend storing your memory.
A closer look at the memory layer
Relational, vector, and graph data in a single ACID-compliant, globally distributed store.
Query across all memory at once
Semantic recall, graph queries, and recent history in one request, fused into a single ranked, consistent result without client-side coordination.
Atomic writes, zero sync lag
Write execution logs, embeddings, and graph edges in one transaction. Read-your-writes consistency across every memory type with no sync lag between stores.
Time-travel any moment
Read an agent's full memory state as it existed at any past moment (logs, vectors, and graph together) for debugging, audit, and reproducibility.
Atomic erasure
Delete an agent and purge all associated memories, vectors, and graph relationships in a single transaction.
Pinned where your agents run
Pin agent memory to a home region and serve it from the nearest data plane, keeping read and write latency low across single regions, multi-region geographies, or globally.
API & SDK first
Manage agents, write memory, and run retrieval through a clean API and typed SDKs.
Who's building Jennah
Jennah is built by NightBlue, a Tokyo-based company founded by Chew Esmero.
Chew Esmero
Founder, NightBlue | CTO, Alphaus
Chew has spent over two decades building low-level systems, distributed databases, and cloud data platforms. As CTO of Alphaus, he directs enterprise cloud infrastructure. He founded NightBlue to build Jennah as a unified backend for stateful agents.
Become a design partner
We are partnering with teams running agents in production: hands-on access, direct founder support, and input on the roadmap in exchange for candid feedback. Jennah is currently in early access, so pricing is bespoke while we build with our initial cohort. To ask a question or request a demo, use the form below.
Prefer email? Reach us at info@nightblue.io.