The simplest local-first
feature store for Python.
Define features once. Train models without data leakage. Serve features in milliseconds. No Kubernetes. No Redis. No Cloud. Just Python.
Everything you need, nothing you don't
BoxLink gives you a production-grade feature store that runs entirely locally — no cloud dependency, no DevOps overhead.
Zero Infrastructure
No Kubernetes. No Redis. No Docker. BoxLink runs entirely on your local machine with SQLite and DuckDB embedded.
DuckDB Offline Joins
Point-in-time correct historical feature retrieval powered by DuckDB. No data leakage, ever.
Sub-millisecond Serving
SQLite-powered online feature store delivers features in under 1ms. Fast enough for real-time inference.
Feature Versioning
Version your feature views, track lineage, and roll back safely. Your features evolve without breaking pipelines.
Define Once, Use Everywhere
Write feature logic once. Reuse consistently across training, evaluation, and production serving.
Pure Python
Install with a single pip command. No config files, no YAML sprawl, no cloud accounts required.
downloadsWorks with your whole stack
BoxLink integrates natively with the Python ML ecosystem you already use.
Why BoxLink for Feature Engineering
Traditional feature engineering fails during model deployment. BoxLink bridges the gap.
Performance Benchmark
How BoxLink stack measures up against cloud-heavy traditional feature store infrastructure.
BoxLink (Local-first)
Local SQLite/DuckDB retrieval times
Traditional Store (Feast)
Cloud infrastructure retrieval latency
Architecture
DuckDB powers point-in-time offline historical joins, and SQLite delivers sub-millisecond serving online. Everything runs on your local machine — no cloud required.
BoxLink vs. Feast
A comparison of features, infrastructure requirements, and local development velocity.
| Capability | BoxLink | Feast |
|---|---|---|
| Local-first / Zero-infrastructure | Yes | Complex |
| Kubernetes Required | No | Often |
| Redis Required | No | Usually |
| DuckDB (Offline Joins) | Yes | No |
| SQLite (Online Serving) | Yes | No |
| Point-in-Time Joins | Yes | Yes |
| Feature Versioning | Yes | Yes |
| Zero Setup | Yes | No |
Designed for local speed. Ready for enterprise scale.
Start instantly on your laptop using the open-source library. When your workloads grow, transition smoothly to BoxLink Cloud for collaborative, managed serverless compute.
Open Source
- Local-first embedded architecture
- Zero infrastructure or cloud setup
- Free forever under MIT License
BoxLink Cloud
- Centralized governance & security
- Automated sync & serverless scales
- Only pay for what you query
2-Minute Quick Example
Register features, perform historical temporal joins, and materialize to SQLite for sub-millisecond serving.
from boxlink import Entity, Feature, FeatureView, BoxLinkClient
client = BoxLinkClient()
# Identify entity
customer = Entity(name="customer_id", value_type="INT64")
client.register_entity(customer)
# Define feature view mapping to local Parquet file
fv = FeatureView(
name="customer_stats",
entities=[customer],
features=[
Feature(name="balance", dtype="double"),
Feature(name="active_days", dtype="int64")
],
source_path="customer_data.parquet",
timestamp_field="timestamp"
)
client.register_feature_view(fv)Complete ML Examples
Browse full end-to-end setups implementing BoxLink in standard machine learning pipelines.
Customer Churn
End-to-end model training, temporal offline evaluation, and low-latency online inference serving.
Fraud Detection
Dynamic query-time On-Demand transformations, sub-millisecond lookups, and feature pipelines.
House Price Prediction
Aggregating zip-code statistics and spatial-temporal features using correct point-in-time joins.
Recommendation System
Version fallback handling, dynamic entity lookups, and flexible recommendation-oriented APIs.
Time Series Forecasting
Generating target lag features, shifting window features, and correct temporal data alignments.
Performance Benchmarks
Testing cache hit vs. miss speedups, and analyzing Pandas vs. Polars extraction timings.
Loved by ML Engineers
From startups to research labs — teams building real ML systems trust BoxLink.
“BoxLink completely replaced our Feast setup. We went from a 45-minute environment setup to `pip install boxlink` and done. Point-in-time joins work out of the box without any Spark dependency.”
“I was skeptical about a local-first feature store but BoxLink proved me wrong. The DuckDB-powered historical joins are faster than anything I ran on Spark clusters. No cloud bills either.”
“We use BoxLink across our entire team's laptops for development and the same codebase deploys to Lambda for serving. The zero-infrastructure philosophy is exactly what a lean ML team needs.”
Latest From The Blog
Deep dives into local-first feature engineering, real-time machine learning, and data pipelines.
Building Local-First ML Pipelines with DuckDB & SQLite
How to utilize embedded databases to solve offline feature generation and online serving without standing up heavy network infrastructure.
Understanding Point-in-Time Joins and Data Leakage
A detailed breakdown of temporal joins, why traditional joins cause target leakage, and how BoxLink automates AS-OF joins.
From Laptop to Production: Deploying BoxLink in Serverless
Learn how BoxLink's zero-dependency design makes it the perfect fit for AWS Lambda, Google Cloud Run, and edge runtimes.