Job Description

We are hiring an LLM Ops Engineer to join our AI Research team, a highly technical group working on cutting-edge advancements in the AI industry. The team focuses on building scalable, production-grade LLM systems, fine-tuning strategies, evaluation frameworks, and next-generation deployment architectures.

This role requires hands-on experience operating LLMs beyond easy API integration. The ideal candidate understands the architectural, operational, and evaluation complexities that differentiate LLMOps from traditional MLOps.

Responsibilities:-

- Manage the end-to-end lifecycle of LLMs: registry, packaging, versioning, deployment, monitoring, and rollback.
- Deploy and operate self-hosted / open-source LLMs (not limited to OpenAI API usage).
- Design and manage scalable inference infrastructure, including GPU-aware deployments.
- Implement CI/CD pipelines for LLM deployment and continuous evaluation.
- Monitor system performance including latency, throughput, token usage, cost, drift (model and data), and hallucinations.
- Ensure secure,



compliant, and resilient cloud-based model deployments.
- Collaborate with research and engineering for deployments.

Skills:-

- Strong hands-on experience with LLM handling, hosting, and operationalization.
- Clear understanding of how LLMOps differs from traditional MLOps (prompt management, non-deterministic outputs, semantic evaluation, token economics, guardrails etc.).
- Experience with Kubernetes, Docker, and containerized deployments.
- Cloud expertise (AWS / Azure / GCP) including compute, storage, IAM, networking, and monitoring.
- Experience building scalable inference and model-serving architectures.
- Familiarity with tools such as MLflow, Kubeflow etc. (good to have).
- Understanding of vector databases, RAG systems, and evaluation frameworks (preferred).
- Knowledge of GenAI security considerations (prompt injection, data leakage prevention).

📌 LLM Ops Engineer (Gurugram)
🏢 Wns Global Services
📍 Gurugram