Tech Lead, LLM

This role requires someone to ship code & lead from the front by acting as a Player/Coach: Architect the system and mentor the team, but spend significant time hands-on in the codebase (Python/PyTorch). You will own the core chat loop: Optimize context windows, memory/RAG retrieval, and inference latency to ensure a seamless, real-time experience. You will also drive our strategy for SFT (Supervised Fine-Tuning) and RLHF/DPO (Preference Optimization). You will be tasked to build the immune system of the platform by designing and training custom classifiers to detect and filter non-consensual or illegal content within an explicit environment.