Senior Software Engineer, RAG and Agentic AI

NVIDIA NVIDIA · Semiconductors · Pune, India

Senior Software Engineer to join the AI Blueprints team to build multimodal, scalable, production-grade reference RAG solutions using Agentic AI and NVIDIA Nemotron models. Develop orchestration layers to dynamically interact with proprietary unstructured and structured data sources. The role involves planning, building, and refining RAG workflows, designing and implementing AI agents for reasoning, planning, and multi-step execution, running POCs, hardening patterns, building and deploying end-to-end RAG pipelines using microservices architecture, and driving continuous improvement through evaluation and performance analysis.

What you'd actually do

  1. Plan, build and refine a GPU-accelerated, scalable, configurable Retrieval Augmented Generation (RAG) workflow and optimize it for accuracy, relevance, grounding and performance.
  2. Design and implement AI agents to enhance RAG pipeline which are capable of reasoning, planning, multi-step execution, and collaboration across tools and services
  3. Run fast, high-quality POCs on emerging agent and RAG architectures; harden successful patterns into generalized, reusable implementations and integrate them as part of production software.
  4. Build and deploy a disaggregated, end-to-end RAG pipeline using on-prem microservices architecture, orchestrating complex, multi-service deployments from local Docker environments to enterprise-scale Kubernetes clusters.
  5. Drive the continuous improvement of the pipelines by rigorously evaluating system accuracy, characterizing performance metrics across components, analyzing the data and recommending actionable strategic enhancements.

Skills

Required

  • Python
  • AI applications
  • LLM-powered AI applications
  • RAG
  • Agentic AI workflows
  • LLM design patterns
  • tool calling
  • prompt engineering
  • structured outputs
  • reasoning
  • agent frameworks
  • orchestration systems
  • microservices
  • Docker
  • Helm
  • Kubernetes
  • end-to-end software lifecycle
  • release packaging
  • CI/CD pipelines

Nice to have

  • multi-agent systems
  • workflow orchestration engines
  • evaluation frameworks
  • MLOps pipelines
  • AI observability tooling
  • deploying AI models on data center, cloud, and embedded systems
  • AI coding agents

What the JD emphasized

  • 8+ years of professional software engineering experience, with deep expertise in Python, and AI applications.
  • Hands-on experience building and deploying LLM-powered AI applications or RAG or Agentic AI workflows.
  • Experience with agent frameworks or orchestration systems such as LangGraph, LangChain, OpenAI Agents SDK, or similar.

Other signals

  • building multimodal, scalable, production-grade reference RAG solutions
  • develop orchestration layers to dynamically interact with proprietary unstructured and structured data sources
  • design and implement AI agents to enhance RAG pipeline which are capable of reasoning, planning, multi-step execution, and collaboration across tools and services