Now in development

The Memory-Native
Programming Language

A language where nothing is forgotten. Temporal queries, semantic search, and AI-native types - built into the language itself.

Fongbe "E flin nu" - It remembers things
example.flin
// Define an entity - it's automatically persisted
entity User {
    name: text
    email: text
    bio: semantic text    // Auto-embedded for AI search
    joined: time = now
}

// Create and save
user = User { name: "Juste", email: "[email protected]" }
save user

// Time travel - access any past state
old_name = user.name @ yesterday
prev_bio = user.bio @ -1     // Previous version

// Natural language queries
results = ask "active users who joined this month"
Features

Built for how we actually work

Every feature you'd spend months building is already there. Focus on your product, not your infrastructure.

Temporal-Native

Every value automatically maintains its complete history. Time travel through your data with the @ operator.

old = user.name @ yesterday

AI-Native Types

semantic text automatically generates embeddings. Similarity search is a native operation, not a library.

bio: semantic text

Intent-Based Queries

Ask questions in natural language. No SQL, no query builders. Your intent becomes the query.

ask "users from Paris"

Zero Configuration

No database setup. No ORM. No connection strings. Define entities and they just work.

entity Todo { title: text }

Offline-First Sync

Built-in CRDT support for conflict-free replication. Works offline, syncs when connected.

// Auto-synced, conflict-free

High Performance

Compiles to LLVM and WebAssembly. Rust-level performance with Python-level ergonomics.

flin build --target wasm
Compare

The old way vs. the FLIN way

What used to take hundreds of lines now takes one.

:(

Traditional Approach

# Python + SQLAlchemy + Pinecone + 200 lines...

from sqlalchemy import create_engine
from sentence_transformers import Model
import pinecone

# Setup database
engine = create_engine('postgresql://...')

# Setup embeddings
model = SentenceTransformer('all-MiniLM-L6')

# Setup vector DB
pinecone.init(api_key='...')

# ... 150 more lines of boilerplate ...
:)

The FLIN Way

// That's it. Really.

entity Document {
    title: text
    content: semantic text
}

// Semantic search - one line
results = search "machine learning" in Document

// Or just ask
answer = ask "What is ML?"
10x
Less code to write
0
Configuration files
oo
History preserved
1
Language to learn
Get Started

Install in seconds

One command. No dependencies. Start building immediately.

$ curl -fsSL https://flin.dev/install.sh | sh

Or install with Cargo: cargo install flin-cli

ZeroSuite Ecosystem

Part of ZeroSuite

FLIN integrates seamlessly with the ZeroSuite family of developer tools.

Ready to remember everything?

Join the FLIN community and build applications where nothing is ever lost.