Lead LLM Engineer

We are looking for a Lead LLM Engineer to drive the design, architecture, and deployment of production-grade Generative AI systems. This is a highly hands-on role (approximately 70% Individual Contributor, 30% Technical Leadershipwhere you will architect scalable AI platforms while mentoring engineers and defining technical direction. You will lead the development of enterprise-grade LLM applications, Agentic AI workflows, Retrieval-Augmented Generation (RAG) systems, and multimodal AI solutions that power real-world products at scale.

Responsibilities:

• Design, build, and deploy production-grade LLM applications using state-of-the-art foundation models and agentic AI frameworks.
• Architect scalable Retrieval-Augmented Generation (RAG) and GraphRAG systems with hybrid retrieval, reranking, and evaluation pipelines.
• Lead the development of multi-agent AI workflows using frameworks such as LangGraph, AutoGen, CrewAI, or similar.
• Fine-tune and optimize LLMs/SLMs using LoRA, QLoRA, PEFT, distillation, quantization, and inference optimization techniques.
• Build robust evaluation frameworks covering hallucination detection, factuality, latency, cost, and response quality.
• Develop multimodal AI solutions combining text, images, documents, and structured enterprise data.
• Collaborate with Product, Engineering, and Design teams to integrate AI capabilities into production products.
• Mentor junior engineers, conduct architecture reviews, establish engineering best practices, and provide technical leadership across the AI team.
• Drive AI platform architecture, observability, experimentation, and MLOps best practices.
• Stay current with advancements in LLMs, Agentic AI, reasoning models, multimodal AI, and enterprise AI infrastructure.

Requirements:

• 5-8 years of experience in Machine Learning, NLP, Deep Learning, or Applied AI.
• 2+ years of hands-on experience building production LLM or Generative AI applications.
• Proven experience leading technical initiatives or mentoring engineering teams.

Technical Skills:

• Strong expertise in Large Language Models (LLMs), Transformer architectures, embeddings, and prompt engineering.
• Hands-on experience designing production RAG/GraphRAG systems.
• Experience building Agentic AI workflows using LangGraph, AutoGen, CrewAI, OpenAI Agents SDK, or equivalent.
• Strong knowledge of vector databases such as Pinecone, FAISS, Qdrant, ChromaDB, Weaviate, or Milvus.
• Experience fine-tuning LLMs using LoRA, QLoRA, PEFT, instruction tuning, or supervised fine-tuning.
• Strong Python programming skills with PyTorch and Hugging Face Transformers.
• Experience building scalable backend services using FastAPI, Flask, or similar frameworks.
• Familiarity with cloud platforms such as AWS, Azure, or GCP, including AI services like Bedrock, SageMaker, Vertex AI, or Azure OpenAI.
• Experience with Docker, Kubernetes, CI/CD, MLflow, LangSmith, observability, and model monitoring.
• Strong understanding of inference optimization, quantization, batching, caching, and GPU utilization.

Nice to Have:

• Experience with multimodal AI (Vision-Language Models, OCR, and Document AI).
• Knowledge of Knowledge Graphs or GraphRAG architectures.
• Experience with distributed training and large-scale model serving (vLLM, TensorRT-LLM, Triton).
• Publications in AI conferences (NeurIPS, ICML, ACL, EMNLP, CVPR, ICCV) or contributions to open-source AI projects.
• Experience building AI evaluation frameworks, safety guardrails, and hallucination mitigation systems.

Expectation:

• You enjoy solving complex AI problems and building production-grade systems.
• You thrive in fast-paced environments and take ownership from architecture to deployment.
• You enjoy mentoring engineers while remaining deeply hands-on.
• You are passionate about Agentic AI, LLMs, and building the next generation of enterprise AI platforms.
• You want to shape AI products used at scale and work on cutting-edge Generative AI technologies.

Skills

GraphRAG, Hugging Face, LLM, PyTorch, Python, RAG, Transformers, generative AI, large language model, large language models, retrieval augmented generation