Celeris Home
MongoDB-compatible Cassandra-compatible App-scoped auto-access

NoSQL without the glue code.

Define MongoDB or Cassandra as a component. If your service and NoSQL live in the same Celeris Application, access is granted automatically—standard drivers work out of the box.

Create NoSQL component See performance tuning
Application: analytics-platform
api Service
Binder Identity
NoSQL Components
users
events
Connected (mongo/cql)

No static secrets. No manual grants.

Explore

Choose your model

Two first-class NoSQL experiences. Pick what fits your data.

users
sessions TTL: 24h
audit_logs
User profiles, content, nested JSON
Rapid iteration, flexible fields
Secondary indexes + aggregation

App-scoped auto-access

Same application = instant connectivity. Standard drivers work with zero configuration.

Application: analytics-platform
api
Binder
NoSQL

Credentials are ephemeral and scoped.

MongoDB Driver
// Connection string from env
const uri = process.env.MONGO_URI;
const client = new MongoClient(uri);
// That's it. Celeris handles the rest.
await client.connect();
CQL Driver
// Contact points from env
const client = new Client({
contactPoints: [process.env.CASSANDRA_HOST],
localDataCenter: 'datacenter1'
});
// Credentials injected automatically
await client.connect();

Data modeling & performance

Indexes matter. Partition keys prevent hotspots. We help you get it right.

Indexes & query shapes

MongoDB: Design for your queries

Query
users
Explain Plan
Stage: IXSCAN
Index: users_email_status

Tip: Compound indexes should match your query patterns. The order of fields matters—put equality filters first, then range filters.

Partition keys & clustering

Cassandra: Design for distribution

(tenant_id) partition key
Hot partition detected
Consider composite key

Tip: Choose partition keys that distribute data evenly. Add a time bucket (e.g., day) to high-cardinality keys to prevent hotspots.

Observability & operations

Real-time health, hotspots, slow queries, and backups—all in one control room.

NoSQL Cluster
24.3K ops/sec
4.2ms p95 latency
847GB storage
2 hot partitions
7 slow queries
3/3 replicas healthy
2h ago last backup
94% index hit rate
Hot Partitions
tenant_1234 45% of traffic
tenant_5678 23% of traffic
Suggested: Add day_bucket to partition key
Demo: Spot hotspot → AI suggests partition key change → approve → apply

AI-first agent

Schema-aware, index-aware, partition-aware. Not a chatbot—a workflow engine.

Ask Celeris

Data Path Analysis

Ingest
Partition
Replica

Plain-English Cause

Write spikes are caused by a batch job running at 2:00 AM UTC that inserts ~50K events/minute. The current partition key (tenant_id) concentrates 45% of this traffic on a single partition.

Demo: Ask → diagnose → index/partition suggestion → run policy tests → approve → apply

JIT access + governance

Time-bound access, approval workflows, and complete audit trails. Enterprise-grade by default.

Access Layers

View data
JIT access • Scoped to collection/keyspace
Run queries
JIT access • Read-only or read-write
Change schema/index/keys
Approval required • Audit logged
Break-glass

Emergency access with heavy audit. Page oncall automatically.

JIT Access Request

Active session

Approved
Requester sarah@company.com
Scope users collection
Permission Read-only
Reason Debug user sync issue
Time remaining 47:32
Auto-expires at 3:15 PM
MongoDB Scope
  • db/collection/fields
  • Query pattern limits
  • Document count caps
Cassandra Scope
  • keyspace/table
  • Partition key ranges
  • Consistency level limits
Audit Trail
14:32 ✓ Access granted to sarah@company.com
14:30 ● Approved by mike@company.com
14:28 ○ Request submitted (JIRA-1234)
Agent-proposed access

AI can suggest access grants based on context. Humans always approve.

Marketplace integrations

Third-party integrations for CDC, analytics, search, and observability—all policy-aware.

NoSQL
Core
Stream CDC
Event Bridge
Warehouse
Object Store
Full-text Search
Vector Search
Metrics
Secrets
Audit Logs
Sync

CDC / Streaming

Real-time change capture

  • • Stream to Kafka, PubSub, Kinesis
  • • Event-driven architectures
  • • Policy-aware filtering
One-click setup

Analytics Export

Data warehouse integration

  • • Export to data warehouse
  • • Object storage snapshots
  • • Scheduled exports
GDPR-aware exports

Observability

Metrics, traces, alerts

  • • Custom metrics pipelines
  • • Distributed tracing
  • • Alert integrations
Pre-built dashboards
Demo: Enable CDC integration → events flow → observe pipeline
Browse Marketplace

NoSQL workbench

A crafted instrument panel for query exploration, schema management, and operations.

Mongo Shell
Interactive MongoDB query console
CQL Explorer
Cassandra CQL query interface
Browser
Collection & table browser with usage hints
Indexes
Index management & recommendations
TTL / Retention
Configure document/row expiration
Partitions
Partition key & clustering configuration
Backups
Backup schedules & point-in-time restore
Migrations
Schema migration tracking & rollback
Preview Env
Isolated data strategies for previews
query.js
// Find users by status with index
db.users.find({
"status": "active",
"created_at": {
$gte: ISODate("2026-01-01")
}
}).sort({ "created_at": -1 })
.limit(100)

Preview-safe data strategies

TTL sandboxes, seeded datasets, isolated namespaces for preview environments

Isolate + Seed

Per-preview isolated collection/keyspace

  • • Auto-created preview_abc123_users
  • • Seed hooks populate test data on create
  • • TTL auto-cleanup after preview closes

Namespace Branch

MongoDB collection namespace branching

  • • Branch from production schema
  • • Copy-on-write semantics (where supported)
  • • Merge or discard on PR close

Ship with NoSQL that's fast, governed, and AI-tuned.

MongoDB and Cassandra compatibility. App-scoped auto-access. Enterprise-grade governance. AI-powered optimization.

Create NoSQL component Browse marketplace Talk to sales
MongoDB-compatible Cassandra-compatible App-scoped auto-access AI-powered tuning