Lead NLP Engineer

🚀 Looking for a cool Lead NLP Engineer

Tasks: 🔹 Optimizing LLM inference in production: SGLang, vLLM, TensorRT-LLM, KV/prefix caching, batching, quantization, speculative decoding 🔹 Distributed inference of models up to 1T+ parameters (multi-GPU, multi-node) 🔹 Benchmarking and introducing new GPU hardware into production 🔹 Technical leadership of NLP and CV teams 🔹 Training and fine-tuning LLM, post-training (RLHF, DPO) 🔹 Working with validation and dataset preparation teams

Expectations: ✅ Deep experience in optimizing LLM inference in production ✅ Experience in distributed inference/training: MoE, tensor/expert/pipeline parallelism ✅ Understanding of KV cache, attention kernels, quantization, GPU profiling ✅ Experience in training and fine-tuning LLM, post-training ✅ Technical leadership + hands-on code ✅ PyTorch, transformers ✅ Experience in AI startups ✅ Backend (Python, Go, C#) — advantage ✅ English or Russian — advanced

Bonus points: ➕ CUDA/Triton kernels ➕ Computer vision ➕ Multimodal LLM

Conditions: 🕐 Remote 📈 Path to Head of ML

📩 Resume: recruiter contact