Hi HN! I built ALEC because I was frustrated with how much bandwidth IoT sensors waste transmitting predictable data.
The core idea: instead of sending raw values, ALEC learns sensor patterns and transmits only the delta from its prediction. After a warm-up period, compression ratios reach 90%+ on typical sensor data.
Key features:
- Evolving shared context between encoder/decoder
- Priority classification (P1-P5) for critical alerts
- Preload system for optimal compression from byte one
- Written in Rust, ~4KB encoder footprint
Use cases: satellite IoT (where every byte costs $$$), remote sensors on battery, agricultural monitoring.
The codec is AGPL-3.0 with commercial licenses available. Happy to answer any questions about the implementation or the information theory behind it!
Hi HN! I built ALEC because I was frustrated with how much bandwidth IoT sensors waste transmitting predictable data.
The core idea: instead of sending raw values, ALEC learns sensor patterns and transmits only the delta from its prediction. After a warm-up period, compression ratios reach 90%+ on typical sensor data.
Key features: - Evolving shared context between encoder/decoder - Priority classification (P1-P5) for critical alerts - Preload system for optimal compression from byte one - Written in Rust, ~4KB encoder footprint
Use cases: satellite IoT (where every byte costs $$$), remote sensors on battery, agricultural monitoring.
The codec is AGPL-3.0 with commercial licenses available. Happy to answer any questions about the implementation or the information theory behind it!