LLMOps Developers - Pre-Req Approval ID: PA26INDAINQ3037

Overview

LLMOps Engineer role based in Hyderabad, focused on maintaining CI/CD and production operations for LLM prompt deployments, vector indexes, and agent configurations.

What you'll do

• Maintain automated CI/CD pipelines for model prompt deployments and agent graph configurations.
• Monitor production model endpoints for latency, token usage, drift, and accuracy issues.
• Administer, re-index, and optimize vector databases (Milvus, Qdrant, Pinecone).
• Build observability dashboards for token costs, rate limits, and model utilization.
• Configure tracing platforms (LangSmith, Phoenix, Arize) to capture agent decisions and tool execution history.
• Support regression test suites before foundation model or system configuration updates.

What you'll need

• 4–7 years in software engineering, DevOps, or ML platform maintenance.
• 2+ years dedicated to LLM deployment operations.
• Hands-on experience with LLM observability, tracing, and evaluation tools (LangSmith, LangFuse, Arize, TruLens).
• Knowledge of vector database management and indexing parameters (HNSW, IVFFlat) and payload filtering optimization.
• Strong Python skills and experience with REST APIs, FastAPI, and containerized deployments.
• Understanding of version control strategies for prompt engineering and vector database migrations.

Details

• Location: Hyderabad, India.
• Work mode: Hybrid.