Description

Build and maintain LLM-based application components including agent loops, RAG pipelines, and data parsing systems.

This role is on-site.

Responsibilities

• Develop and maintain LLM harnesses for agent loops, tool calling, context construction, and failure handling using LangChain, LlamaIndex, or LangGraph.
• Design, version, and evaluate prompts against structured datasets to measure quality deltas and maintain clean prompt libraries.
• Build and operate RAG pipelines involving chunking, indexing, retrieval, reranking, and response handling.
• Generate embeddings and manage vector indexes (OpenSearch, Pinecone, pgvector) to optimize retrieval quality and cost.
• Create data extraction and parsing pipelines for documents and customer datasets to prepare clean, labeled data for AI work.

Required Skills

• 3+ years of AI or data science experience with hands-on LLM application development.
• Strong Python proficiency with the standard data science stack: pandas, numpy, scikit-learn.
• Experience building LLM harnesses against APIs like Anthropic, OpenAI, or Bedrock.
• Working knowledge of at least one orchestration framework (LangChain, LlamaIndex, LangGraph) and vector databases.
• Ability to write clean, tested, production-ready code using Git and CI/CD practices.
• Experience with traditional ML tasks such as classification, clustering, and lightweight fine-tuning.
• Strong analytical mindset for data inspection, error analysis, and iterative experimentation.

Preferred Skills

• Hands-on experience with agentic frameworks (LangGraph, Claude Agent SDK, OpenAI Agents) in production.
• Built structured eval harnesses including golden sets, regression suites, and LLM-as-judge patterns.
• Hands-on AWS experience with Bedrock, SageMaker, and OpenSearch.

Referral Bonus: ₹50,000

Key Skills

Education

Bachelor's or Master's degrees