Data Scientist R&D – Speech Intelligence at ToumAI is a unique opportunity to drive cutting‑edge voice AI research. You will design, train, and optimize speech models for multilingual, low‑resource languages, ensuring high accuracy, low latency, and efficient on‑device deployment. Working closely with platform and ML engineers, you will build end‑to‑end pipelines, from data collection and annotation to evaluation and production integration. The role demands strong expertise in ASR, diarization, emotion recognition, and deep learning frameworks like PyTorch, as well as a passion for low‑latency inference and multilingual speech. Join a dynamic team focused on making every voice understandable across cultures and languages.