The role focuses on building and shipping production LLM and GenAI systems: retrieval-augmented generation pipelines, agentic workflows, fine-tuning, and the evaluation infrastructure required to prove those systems actually work at scale.

This is a hands-on engineering position on a well-funded AI team — expect to own significant components of the platform end-to-end, from model selection and prompt architecture to latency, cost, and reliability in production.

Key Responsibilities

• Design and build RAG pipelines using LangChain, LlamaIndex, or custom architectures, including chunking strategies, hybrid retrieval, and reranking
• Develop agentic workflows with tool calling, planning, and multi-step reasoning — deployed behind robust APIs with strict latency and cost budgets
• Run parameter-efficient fine-tuning (LoRA, QLoRA, DPO) on domain datasets; manage experiment tracking, dataset versioning, and model registries
• Build systematic LLM evaluation frameworks: golden datasets, LLM-as-judge pipelines, offline benchmarking, and pre-deployment regression testing
• Own vector database infrastructure (Pinecone, Weaviate, pgvector, or Qdrant) including embedding refresh strategies and index performance tuning
• Optimize inference performance: quantization, batching, caching, and serving via vLLM, TGI, or managed APIs
• Collaborate with product and applied science teams to translate ambiguous business problems into scoped, shippable LLM features

What We Are Looking For

• 3–6 years of software engineering experience, including at least 1–2 years shipping LLM/GenAI systems to production
• Strong Python engineering skills: async programming, API design, testing, and working with frameworks like FastAPI
• Hands-on experience with at least one major cloud (AWS, GCP, or Azure) and containerized deployment (Docker, Kubernetes)
• Practical knowledge of transformers, embeddings, tokenization, and the tradeoffs between hosted APIs and self-hosted open models (Llama, Mistral, Qwen)
• Experience operating vector databases and semantic search systems under real traffic
• BS/MS in Computer Science, related field, or equivalent practical experience
• Bonus: Experience with RLHF pipelines, multimodal models, structured output decoding (JSON mode, function calling at scale), or contributions to open-source LLM tooling

Are you looking for remote jobs near your area? At Yulys, thousands of employers are looking for exceptional talent like yours. Find a perfect job now.