Skip to content
WeftKitBeta
Limited Beta — Now Accepting Requests

The Enterprise Database Framework
Shipped as Docker.

Seventeen specialized database engines — relational, vector, key-value, document, graph, in-memory, and more — sharing one crash-safe kernel. Distributed exclusively as Docker images. Free to pull and run. Currently in limited beta.

View Benchmarks
Docker-nativeSub-microsecond latencyPull and runNo license key

17 Database Engines

SQL, Vector, KV, Document, Graph, In-Memory, TimeSeries, and more — all on one kernel.

Enterprise Security

Multi-tenant isolation, encryption at rest, RBAC, audit logging, and zero-trust architecture.

Pull and Run

One `docker pull` per engine. No source build, no compiler toolchain, no license key — just pull, run, and connect.

~116 nsp99 latencyIn-Memory reads
8DatabaseEngines
1M+ConnectionsPool Manager
~23 nsBuffer pool pinEngine kernel
14LanguageBindings
Database Modules

Eight Specialized Engines. One Unified Platform.

Each database is purpose-built for its data model — optimized algorithms, native query languages, and industry-standard wire protocols. Share a single storage kernel. Include only what you need.

WeftKitRel

Relational SQL Database

PostgreSQLSQLite
~1.6 μs

simple SQL parse latency

Full-featured embedded SQL database with a cost-based optimizer, MVCC, B+Tree indexes, and vectorized analytics. Superior to SQLite for OLTP workloads.

B+TreeWALMVCCCost-Based Optimizer
Learn more

WeftKitVec

Vector Similarity Search

PineconepgvectorFAISS
~234 μs

flat search (100 vectors, 128-dim)

Disk-resident approximate nearest neighbor search for RAG pipelines, semantic search, and recommendation systems. Handles billion-point datasets on a single machine.

HNSWDiskANNIVF-PQScalar Quantization
Learn more

WeftKitKV

Key-Value Store

RocksDBLMDBDynamoDB
~204 ns

p99 point get latency

High-performance embedded key-value store with dual LSM-tree/B+Tree engine, TTL, merge operators, bloom filters, and column families.

LSM-TreeB+TreeBloom FiltersMerge Operators
Learn more

WeftKitDoc

Document Database

MongoDBCouchDB
~3.1 μs

single document insert

Schema-flexible document database for JSON/BSON documents with secondary indexing, aggregation pipelines, and change streams. MongoDB-compatible query subset.

BSON StorageULID Document IDsAggregation PipelineChange Streams
Learn more

WeftKitGraph

Graph Database

Neo4jAmazon Neptune
~15 μs

BFS 100-vertex traversal

Native property graph database with Cypher-like query language, BFS/DFS traversals, shortest path algorithms, and CSR/CSC storage format.

CSR/CSC StorageCypher-like QLBFS/DFSPageRank
Learn more

WeftKitMem

In-Memory Cache

RedisMemcached
~116 ns

p99 point read (single-process)

Ultra-fast in-memory cache with lock-free hash maps, configurable eviction policies, optional persistence snapshots, pub/sub, and atomic counters.

Lock-free Hash MapsLRU/LFU/ARCPub/SubAtomic Counters
Learn more

WeftKitMD

Markdown Content Database

Custom (no direct equivalent)
~18.6 μs

BM25 search (1K documents)

Purpose-built Markdown database with AST parsing, frontmatter indexing, full-text search, version history, cross-document linking, and rendering.

Markdown ASTFrontmatter IndexingFull-Text SearchVersion History
Learn more

WeftKitFile

File & Object Storage

S3MinIO
~50 μs

64-byte download latency

Content-addressable file storage with deduplication, streaming I/O, metadata indexes, MIME detection, thumbnail generation, and S3-compatible API.

Content-AddressableDeduplicationStreaming I/OMIME Detection
Learn more
Deployment

One Image. Five Deployment Shapes.

The same Docker image runs unchanged from a single dev container to a Kubernetes StatefulSet. Topology is configuration, not a rebuild.

Single Engine

One engine image on one host. The simplest deployment.

  • One `docker run` command
  • Persistent volume for data
  • Standard wire protocol port
  • Ideal for dev, CI, and internal tools
Client (psql, redis-cli, ...)
weftkit/{engine}:0.1.0
Named Volume: engine-data
Integrations

Speaks Every Language

Every engine image answers a standard wire protocol. Your existing SQL, Redis, Mongo, Neo4j, S3, DynamoDB, and gRPC clients connect with no WeftKit-specific driver.

Python
JS
JavaScript
Java
C#
Go
Swift
Kotlin
Dart
Ruby
Elixir
PHP
PHP
C
Composability

Compose Your Perfect Stack

Include only what you need. Every component is opt-in with zero overhead when not used.

Mobile App

Offline-first mobile app with cloud sync

Component Stack

WeftKitRel
Persistence
Standalone
Estimated memory footprint~8 MB base
Security

Security Built In

Not bolted on. Built in. From the first line of Rust to the last encrypted byte on disk.

Zero-Trust Model

Even embedded databases require authentication. No operation is trusted by default. SecurityContext propagates through every component.

Encryption at Rest

AES-256-GCM with per-page nonces. Envelope encryption with O(1) key rotation without re-encrypting data.

Encryption in Transit

TLS 1.3 mandatory for all network communication. Certificate pinning. Perfect forward secrecy with ECDHE cipher suites.

RBAC Everywhere

Role-based access control from embedded mode through Pool Manager. Per-database-type authorization with fine-grained permissions.

Compliance Ready

SOC2, GDPR, HIPAA compliance helpers. Full audit trail with tamper-evident logging. SBOM generation for every release.

Injection Prevention

SQL, NoSQL, Cypher, command injection, and path traversal protection for every query language and API endpoint.

Compatibility

Drop-In Compatible Wire Protocols

Use your existing tools. Connect with psql, redis-cli, mongosh, or any standard client.

Relational

PostgreSQL v3

psqlJDBCasyncpglibpq
In-Memory

Redis RESP3

redis-cliredis-pylettuce
Document

MongoDB Wire

mongoshPyMongomotor
Graph

Bolt (Neo4j)

neo4j-driverpy2neo
Key-Value

DynamoDB REST

AWS SDKboto3
Vector

gRPC

Custom clientgrpcurl
File Store

S3 REST

AWS CLIboto3rclone
Markdown

REST API

curlhttpieAny HTTP
Limited Beta

Request Beta Access

WeftKit is currently available to a limited number of early adopters. Tell us about your project and we'll provide you with the distribution package and integration documentation tailored to your needs.

  • Pre-built binaries or Docker images for your selected database
  • Integration documentation tailored to your platform and language
  • Direct engineering support during the beta period

We review every request personally. Expect a response within 2 business days.

17Database engines
RustBuilt entirely in
BetaCurrent stage
LimitedAccess model
Beta Access

Ready to build with WeftKit?

Request access to get your personalized distribution package with pre-built binaries and integration documentation.

Contact Us