LOKI (NLP Engine)
Local-First Voice Assistant with Hybrid NLU
99.77%
NER F1-Score
<60ms
Fast-Path Latency
1389
Training Samples
Overview
A privacy-centric voice assistant that performs 100% local inference. It addresses the latency and privacy issues of cloud-based assistants by employing a novel dual-layer intent classification system: utilizing high-speed vector embeddings for common commands and falling back to a local Quantized LLM for complex semantic understanding.
System Architecture
Technical Decisions
| Decision | Trade-off | Outcome |
|---|---|---|
| Hybrid Classifier Strategy | System Complexity vs. Responsiveness | Implemented a 'Fast Path' (Cosine Similarity > 0.6) for instant execution of 90% of commands, reserving the heavy LLM only for complex, novel queries. |
| CRF for NER | Modernity vs. Efficiency | Chose Conditional Random Fields over BERT for Named Entity Recognition to minimize CPU footprint while maintaining 99.77% F1-score on parameter extraction. |
| Threaded Architecture | Dev Overhead vs. UX | Decoupled audio acquisition (VAD) and inference workers from the UI thread, ensuring the application remains responsive during heavy processing. |