Founding AI/ML Research Engineer

This founding technical role defines the core model stack, training strategy, and long-term architecture for a groundbreaking global AI product. You will design and own end-to-end pipelines (data → training → eval → inference), experiment with frontier models (LLaMA, Mistral, Qwen, Claude-compatible), and build scalable inference systems using vLLM, TensorRT-LLM, and DeepSpeed. You’ll implement data systems for high-quality synthetic and real-world data, develop alignment and guardrail strategies, and design evaluation frameworks across performance, robustness, safety, and bias. You’ll own deployment—GPU optimization, latency reduction, and scaling policies—while shaping early product directions and exploring techniques like retrieval-augmented training, mixture-of-experts, distillation, and multi-agent orchestration. This role operates with a founder’s mindset: fast iteration, rigorous engineering, and measurable impact on a global AI product.