|
Описание: |
We are seeking an experienced Machine Learning Engineer specializing in Large Language Models (LLMs), Natural Language Processing (NLP), and AI Agents. In this role, you will design, develop, and deploy cutting-edge language models and agent-based systems that drive innovation and deliver value across products and services.
Key Responsibilities
* Design, fine-tune, and optimize large language models (LLMs) for specific use cases and domains
* Develop and implement prompt engineering techniques to enhance model capabilities and performance
* Create autonomous AI agents capable of performing complex tasks through language understanding and reasoning
* Build retrieval-augmented generation (RAG) systems that combine LLMs with external knowledge sources
* Implement evaluation frameworks to benchmark LLM and agent performance across various metrics
* Optimize models for production deployment with attention to latency, throughput, and cost
* Design conversational flows and dialogue management systems for AI assistants
* Develop techniques to ensure responsible AI deployment, including safety guardrails and bias mitigation
* Stay current with the rapidly evolving NLP/LLM research landscape and implement state-of-the-art techniques
Requirements
* Strong experience with transformer-based models and frameworks (Hugging Face, PyTorch, TensorFlow)
* Practical experience fine-tuning, serving, and optimizing LLMs (e.g., GPT models, Llama, Claude, Mistral)
* Expertise in NLP techniques including prompt engineering, embeddings, and semantic search
* Experience with vector databases and retrieval systems (Pinecone, Weaviate, Faiss, etc.)
* Familiarity with model quantization, distillation, and other optimization techniques
* Experience with LLM orchestration frameworks (LangChain, LlamaIndex, etc.)
* Working knowledge of cloud platforms (AWS, GCP, Azure) for ML model deployment
* Proficiency with containerization technologies (Docker, Kubernetes)
* Strong understanding of LLM evaluation methodologies and metrics
Nice-to-Have skills:
* Experience developing multi-agent systems or agent-based architectures
* Knowledge of reinforcement learning from human feedback (RLHF) techniques
* Familiarity with LLM inference optimization (vLLM, TensorRT, ONNX)
* Experience with tools for LLM observability, evaluation, and debugging
* Experience with multimodal models combining text, vision, and audio
* Knowledge of LLM security, safety, and alignment techniques
Відгукнутись на вакансію |
