Data Scientist
Posted Updated
Summary
Data Scientist (7-10 yrs) in Pune building LLM-powered applications: designing RAG/Graph RAG retrieval pipelines, stateful agentic workflows with LangChain/LangGraph, prompt engineering, and FastAPI integrations, using Python and ML libraries like PyTorch/TensorFlow. Azure, PySpark and Databricks knowledge is a plus.
Primary Skills (Must Have):
- Python Programming: Advanced proficiency, including asynchronous programming, data structures and object-oriented design.
- Large Language Models (LLMs): In-depth understanding and practical experience with LLMs such as OpenAI GPT-3/4, Anthropic Claude, Meta Llama or similar; familiarity with Small Language Models (SLMs).
- RAG Architectures: Knowledge RAG, Graph RAG, hybrid retrieval systems and context-aware retrieval pipelines.
- LangChain & LangGraph: Hands-on experience building, deploying and maintaining applications using RAG, LangChain and LangGraph.
- Graph-Based Workflow Design: Ability to architect, implement and debug complex, stateful, branching conversational flows using LangGraph.
- Agentic Frameworks & Evaluation: Ability to develop workflow automation using an agentic framework, with familiarity with evaluation techniques and tools for agentic applications.
- Prompt Engineering: Expertise in designing, testing and optimizing prompts for generative models.
- API Development & Integration: Experience integrating external APIs, tools, microservices and data sources into AI-driven workflows (FastAPI).
- Machine Learning Libraries: Familiarity with PyTorch, TensorFlow, Hugging Face Transformers or similar.
Secondary Skills (Good to Have):
- Data Engineering: Basic knowledge of PySpark and Databricks is a plus.
- Azure Cloud experience.