LLM / GenAI Engineer

About The Role The role focuses on building, optimizing, and scaling production-grade Generative AI systems, moving far beyond basic API wrappers to engineer robust agentic workflows, custom Retrieval-Augmented Generation (RAG) pipelines, and fine-tuning frameworks. The engineer will collaborate closely with backend, data, and product teams to translate unstructured data into highly accurate, low-latency AI features, ensuring model reliability, guardrails, and cost-efficiency at scale. Key Responsibilities Design and deploy production-grade RAG systems leveraging advanced chunking, hybrid search, and reranking techniques Develop and evaluate agentic workflows using frameworks like LangGraph, CrewAI, or custom orchestration logic Fine-tune open-source LLMs (such as Llama, Mistral) using PEFT techniques like LoRA, QLoRA, and DeepSpeed on domain-specific datasets Build and scale vector database integrations utilizing pgvector, Pinecone, or Qdrant for millisecond-latency semantic retrieval Implement systematic LLM evaluation and observability pipelines using tools like Phoenix, LangSmith, or custom LLM-as-a-judge frameworks Optimize model inference and deployment strategies using vLLM, TensorRT-LLM, or Triton Inference Server to reduce latency and hosting costs What We Are Looking For 3-6 years of professional software engineering experience, with at least 1.5 years dedicated to building and deploying LLM applications in production Proficiency with advanced LLM orchestration tools and frameworks such as LangChain, LlamaIndex, or DSPy Hands-on experience with vector databases, vector indexing strategies, and embedding models Strong Python development skills, including experience with asynchronous programming, FastAPI, and robust unit/integration testing Solid understanding of cloud infrastructure (AWS or GCP) and containerization using Docker and Kubernetes Bonus: Experience with direct GPU optimization, deep learning frameworks like PyTorch, or contributing to open-source GenAI projects Show more Show less