Push raw data.Get structured memory.
RushDB turns JSON, events, and documents into a typed, queryable graph — semantic search, relationships, and analytics for agents and AI apps, in one context layer.
Typical stack
AI context is scattered by default.
Application records live in one database. Session state lives in Redis. Embeddings live in a vector store. Relationships live in code, prompts, or nowhere.
Store context
Application DB
facts, history, metadata
Redis
session state, cache
Vector DB
embeddings, similarity search
Your code owns:
sync · embed · index · join · retry
With RushDB
JSON in. Queryable graph out.
RushDB keeps records, relationships, schema, values, and embeddings together in one queryable context layer.
JSON / CSV
Records
Typed fields
Searchable text
Graph links
Live schema
Analytics
Agents and applications can inspect, search, traverse, update, and calculate metrics over the same grounded structure.
How it works
From raw payloads to queryable context.
Send JSON, CSV, events, documents, or tool output. RushDB creates connected records, tracks discovered properties and values, suggests relationships, and exposes everything through a consistent query API.
json input
push→
graph — auto-linked nodes
vector embeddings
Structured memory layer
Schema on write. Awareness on read.
RushDB does not only remember text. It keeps structure with the data — so AI systems operate against grounded context instead of guessed structure.
01 /Structure on write
Every record you push updates a typed, live schema and links into the graph — at write time, not in a pipeline.
02 /Data that explains itself
Agents query labels, types, and relationships to rebuild full awareness of what is stored. No schema docs, no prompt lore.
03 /Models change. Memory holds.
A new SOTA model ships every few weeks. Awareness is reconstructed from the data, so swapping the model — or the framework, or the MCP client — does not reset your outcomes.
Change the model without rebuilding context.
Durable context layer
Quickstart
Running in minutes.
Install
Add the SDK for TypeScript, Python, or use the REST API directly — no extra dependencies.
Push
Write any JSON object. RushDB parses field types, builds the graph structure, and indexes vectors server-side on the same write.
Recall
Query by meaning, by graph relationship, or both in a single call using the same SearchQuery shape every time.
install
pip install rushdb
push · recall
from rushdb import RushDB
db = RushDB('RUSHDB_API_KEY')
# 1. Push any JSON — no schema, no pipeline
db.records.create(
label='MEMORY',
data={'agent_id': 'agent-42', 'action': 'summarized', 'output': text},
)
# 2. Recall by meaning — semantic search + graph filter in one call
results = db.ai.search({
'labels': ['MEMORY'],
'query': 'what did the agent do?',
'where': {'agent_id': 'agent-42'},
})Use cases
Structured context for real workloads.
The same structured memory layer supports AI agent memory, GraphRAG, applications, sales, analytics, support, and healthcare without hard-coding a new backend for each workflow.
Persistent context
Decisions, tool outputs, and entities survive across runs as reusable records.
Semantic recall
Recall prior work by meaning, scoped to a session, agent, or task.
Shared state
Multiple agents read and write the same durable memory graph.
Features
Push data once. Query it five ways.
Agents can discover what exists, construct valid queries, retrieve by meaning, traverse connections, and calculate metrics without switching backends or maintaining a separate graph pipeline and vector sync job.
Explore all features
Field notes
Structured context built for production teams.
RushDB is built for the context requirements production AI teams actually face: connected records, live schema, semantic retrieval, analytics, and a storage model software can safely inspect.
“In a few minutes, RushDB can go from Docker Compose to Python SDK to nested JSON turning into a real graph. For developer education, that matters: builders can see the data land in the visualizer and Neo4j Aura, not just read about it.”
Fractional DevRel | Video-First Developer Content & Community Growth | ex-Neo4j
Pricing
Start free. Scale as your context grows.
Free includes 100K KU/month. Start is $8/month. Reads are free; writes are usage-based. Full pricing details →
MonthlyAnnual
KU = Knowledge Unit
Free
FREE
Evaluate RushDB without a card
Get started instantly:
- 100K KU / month
For evaluation and early prototypes
- 2 projects
- Full REST API and SDKs
- Self-hosted & BYOC
- Vector & AI search
- Community support
Start
from$8$8 / mo
For production pilots and small apps
Everything in Free, plus:
- 250K KU / month
For solo devs shipping to production
- 2 projects
- 1 team member
- Self-hosted & BYOC
- Community support
Pro
from$24$24 / mo
Perfect for production applications
Everything in Start, plus:
- 1M KU / month
Included before overage
- Overage billing
Beyond 1M KU at $10 / M KU
- 10 projects
- 3 team members
Add more at $10 / seat
- Self-hosted & BYOC
Scale
from$73$73 / mo
For high-volume and data-intensive apps
Everything in Pro, plus:
- Unlimited KU
Usage-based at $8 / M KU
- 100 projects
- 10 team members
Add more at $25 / seat
- Self-hosted & BYOC
- SLA guarantee
- Priority support