Overview

The next part of your journey is right around the corner — with Standard India

Responsibilities

• Ability to work on-site in Bengaluru, India is a requirement of the role. *
• Design and operate the multi-provider model gateway — routing, failover, rate limiting, caching and usage attribution (LiteLLM-style patterns) — across Anthropic, OpenAI and self-hosted open-source models.
• Stand up and operate the MCP/tool registry, AI Gateway — a governed catalog of MCP servers and tools with scoped permissions — and A2A coordination surfaces.
• Set up and operate GPU inference infrastructure on NVIDIA Grace Blackwell (GB300) environments: deploy NVIDIA NIM microservices and open-source serving frameworks such as vLLM, SGLang and NVIDIA Dynamo for distributed serving.
• Manage GPU environments end to end: Kubernetes GPU scheduling (GPU Operator, device plugins, MIG/time-slicing), driver and CUDA stack lifecycle, capacity planning, and throughput/latency/utilization tuning.
• Run the model registry with versioning, lineage, approval workflow and model-risk status for platform-owned models.
• Build LLMOps capabilities: prompt and context versioning, evaluation harnesses, regression testing and token-cost monitoring.
• Stand up LLM observability and evaluation with tools such as Langfuse, Arize Phoenix and RAGAS.
• Integrate guardrails enforcement so governance policy executes at runtime.
• Build MLOps automation for platform-owned predictive models: training pipelines, feature store integration, drift detection and retraining triggers on Databricks.
• Operate model serving and training/evaluation environments on the shared substrate, benchmarking self-hosted models against provider APIs for cost and performance.
• Produce reproducible evidence for independent model validation and mentor junior AI/ML Platform Engineers.

Qualifications

• Typically requires 10+ years of software or ML engineering experience, including significant production ML/AI platform work.
• Expert Python and strong engineering fundamentals: testing, CI/CD and containers.
• Hands-on experience with Anthropic and OpenAI APIs and multi-provider gateway/routing patterns (LiteLLM-style).
• Hands-on GPU inference serving experience with NVIDIA NIM and/or open-source frameworks such as vLLM, SGLang.
• Experience operating GPU infrastructure on Kubernetes: GPU Operator, device plugins, MIG or time-slicing, and utilization/throughput tuning.
• MLflow and model registry experience, plus model serving and monitoring in production.
• LLM evaluation and observability experience with Langfuse, Arize Phoenix, RAGAS or equivalent.
• Experience implementing guardrails or safety controls for generative systems.
• Feature store experience and drift monitoring for predictive models
• Experience deploying on Grace Blackwell-class systems (GB200/GB300 NVL72), NVIDIA AI Enterprise, NVIDIA Dynamo for disaggregated serving, or low-precision formats such as NVFP4/FP8.
• Databricks ML stack depth (Model Serving, Mosaic AI tooling) and Unity Catalog model governance.
• Retrieval engineering with vector stores or GraphRAG (Neo4j).
• Exposure to NIST AI Risk Management Framework or AI TRiSM in a regulated industry.
• Fine-tuning open-source models; GPU FinOps and inference cost optimization.
• Good to have Databricks Machine Learning Professional or Azure AI Engineer certification.
• Adaptability — Seeks information to understand the rationale for and importance of the change.
• Customer Focus — Displays an interest in the customer by trying to understand their concerns and issues; draws on customer insight to help others best meet current and future customer needs.
• Improvement Mindset — Demonstrates curiosity by asking questions regarding current approaches/methods and identifying potential changes.
• A rich benefits package that supports your health, well-being, and financial goals
• An annual incentive bonus plan tied to both individual and organizational performance
• Paid time off including earned leave, sick/casual leave, and public holidays, in accordance with the company’s leave policy
• A supportive, responsive management approach that empowers you to grow
• Opportunities for career advancement through learning, coaching, and leadership enablement
• A culture where your voice matters - your ideas shape our future, and we win as one
• Purpose-driven work that makes a difference for our customers, communities, and each other