AI powered cloud personal data storage.
Schema-based storage, intelligent ingestion, and real-time
processing.
Build smarter data applications with less code.
Download pre-built binaries for your platform or build from source.
M1, M2, M3, M4
x86_64
x86_64
Or build from source:
cargo install --git https://github.com/shiba4life/fold_db --bin datafold_cli
From AI-powered ingestion to distributed queries, FoldDB handles the complexity so you can focus on building.
Drop any JSON and let AI automatically generate schemas, map fields, and structure your data. No manual schema definitions required.
let result = ingestion
.process_json(any_json_data)
.await?;
// Schema auto-generated ✨Ask questions in plain English. AI interprets your intent and returns structured results.
Event-driven architecture with automatic transform execution as data flows through.
Built-in peer discovery and networking. Scale horizontally without infrastructure changes.
Trust-based access control at the field level. Multi-tenant isolation out of the box.
S3-backed storage and DynamoDB support. Deploy to AWS Lambda with zero modifications.
Schemas evolve with your data. Add fields, update validation, all without migrations.
Plugin system for Twitter, Reddit, webhooks, and custom data sources.
Clean, intuitive APIs that get out of your way. Ship faster with less boilerplate.
Just send your JSON data — FoldDB automatically analyzes it, generates an optimized schema, and stores it.
use datafold::{IngestionCore, IngestionConfig};
use serde_json::json;
let config = IngestionConfig::from_env()?;
let ingestion = IngestionCore::new(config)?;
// Drop any JSON — AI handles the rest
let data = json!({
"user": "alice",
"action": "purchase",
"total": 99.99
});
let result = ingestion
.process_json_ingestion(data).await?;Ask questions in plain English and get structured results.
use datafold::DataFoldNode;
let node = DataFoldNode::new_with_defaults().await?;
// Natural language query
let response = node.ai_query(
"Show me all purchases over $50"
).await?;
for item in response.results {
println!("{}: ${}", item.user, item.total);
}First-class serverless support with DynamoDB multi-tenant isolation.
use datafold::lambda::{LambdaConfig, LambdaContext};
let config = LambdaConfig::new(
LambdaStorage::DynamoDb(DynamoDbConfig {
region: "us-east-1".to_string(),
tables: ExplicitTables::from_prefix("MyApp"),
auto_create: true,
}),
);
LambdaContext::init(config).await?;Full TypeScript support with intelligent caching and standardized error handling.
import { schemaClient } from '@datafold/client';
const response = await schemaClient.getSchemas();
if (response.success) {
const schemas = response.data;
schemas.forEach(s => console.log(s.name));
}Install via Cargo or download a pre-built binary.
cargo install datafoldSet your OpenRouter API key for AI features (optional).
export OPENROUTER_API_KEY="sk-..."
Start the HTTP server with the web UI.
datafold_http_server --port 9001
Open localhost:9001 and start ingesting data.
open http://localhost:9001Ingest data, run queries, and manage schemas without leaving the command line.
Verify your FoldDB instance is running and browse available schemas.
-p$ datafold_cli status -p
{
"status": "ok",
"schemas_loaded": 5,
"data_root": "~/.datafold/data"
}
$ datafold_cli schema-list -p
[
"tweets", "contacts", "notes",
"purchases", "bookmarks"
]Feed JSON files directly, or let Smart Folders auto-detect and ingest data from directories.
--all-recommended$ datafold_cli ingest data.json
✓ Schema generated: user_events
✓ 42 records ingested
$ datafold_cli smart-folder-scan ~/Documents
⬡ 12 files detected, 3 schemas recommended
$ datafold_cli smart-folder-ingest ~/Documents --all-recommended
✓ Ingested 3 schemas from ~/DocumentsRun structured queries, full-text search, or ask questions in natural language with AI.
$ datafold_cli query tweets --fields text,author
{ "text": "Hello world", "author": "alice" }
{ "text": "Rust is great", "author": "bob" }
$ datafold_cli search "machine learning"
⬡ 4 results across 2 schemas
$ datafold_cli llm-query "Show me recent purchases over $50"
⬡ Querying with AI...
Found 7 results in schema purchasesLoad, approve, and inspect schemas directly from the CLI.
$ datafold_cli schema-load
✓ Loaded schema: tweets (pending approval)
$ datafold_cli schema-approve tweets
✓ Schema "tweets" approved
$ datafold_cli schema-get tweets -p
{
"name": "tweets",
"fields": ["text", "author", "timestamp"],
"status": "approved"
}Browse the shared schema registry. Click on a schema to view its complete definition including fields and topology structures.
Loading schemas from registry...
Join developers building the next generation of data applications with FoldDB.