Generative AI Engineer

GenAI Engineer

Location: Bengaluru, India (Hybrid)

Experience: 3+ years

We at Mindsprint are looking for a GenAI Engineer with 3+ years of experience to join our team, building and operating production agentic AI systems end-to-end — from architecture through deployment and operation.

Responsibilities

• Design and build multi-agent systems using LangGraph/LangChain, deployed on Azure OpenAI and Databricks

• Own production agentic pipelines end-to-end: architecture, development, deployment (MLflow pyfunc on Databricks), and post-launch debugging

• Build hybrid RAG pipelines (BM25 + dense retrieval + RRF + reranking) for enterprise document and knowledge-base use cases

• Integrate LLM agents with enterprise systems — SAP, Microsoft Graph API (Outlook/mailbox automation), and internal document repositories

• Build and maintain a RAG platform architecture on Databricks serving multiple downstream assistants (Legal, Supply Chain, Contracts)

• Develop classification and validation layers, including YAML-based schema design and LLM-as-Judge evaluation frameworks

• Build internal tooling (FastAPI backends, Streamlit validators, React-based dashboards) to support GenAI operations and observability

• Debug production issues across the stack — LangGraph state/session management, API integration failures, mailbox/data pipeline issues.

Required skills:

• Proficiency in Python, FastAPI, and SQL

• Hands-on experience with LangGraph/LangChain and building real multi-agent orchestration systems (not just IDE-assistant-level GenAI exposure)

• Experience with hybrid RAG architectures: BM25, dense retrieval, RRF, reranking (e.g., FlashRank), and vector databases (e.g., Milvus)

• Experience with PyTorch and model fine-tuning (e.g., Ollama)

• Experience with Azure OpenAI or equivalent LLM API platforms

• Proficiency in Databricks/MLflow for model deployment

• Familiarity with Microsoft Graph API or similar enterprise integration APIs.