On-Device AI Inference

Pack small fine-tuned models into mobile or robot applications for local inference tasks like bar-talk or firefighting support, avoiding cloud-based inference costs.