96.95% HumanEval on a $400 GPU: Frontier-Level Coding Performance on Consumer Hardware
Qwen3.6-35B-A3B scores 96.95% on HumanEval (matching 2025 frontier models) running locally on a $400 GPU at zero cost per token.
Read Case Study →Newsroom
Case studies and notes on shipping AI to hardware. Speech, vision, ranking, and the optimization that makes it fit.
AuraFits: Multimodal AI Advisory Platform for Specialty Retail
Vision-driven product recommendations from 52 live Shopify inventories. One photo, a short conversation, ranked picks from real shelf stock.
Read Case Study →
Semantic-Aware ASR Evaluation System for Edge Devices
On-device ASR evaluation at Fortune Global 50 scale. 90% faster evaluation, 4,000× throughput, deployed to smartphones.
Read Case Study →