We are seeking a skilled Software QA Test Developer to join our test development team passionate about Conversational AI products. This is an outstanding prospect to be part of NVIDIA's legacy of innovation and help develop the future of computing powered by AI. In this role, you will own quality for NVIDIA's voice AI pipelines — end-to-end systems integrating ASR, LLM, and TTS models built on open-source frameworks and NVIDIA NIM microservices. You will compose test strategies, build AI-powered test automation, and drive CI pipeline reliability across multi-model, multi-agent voice systems deployed on cloud, workstation, and edge targets.
What you'll be doing:
Develop and sustain CI/CD pipelines for voice agent projects, automating build, deploy, and test workflows across cloud and GPU deployment profiles.
Develop Python-based test frameworks to validate cascaded voice pipelines (ASR → LLM → TTS), including latency, accuracy, and interruption-handling scenarios.
Track and improve code coverage across Conversational AI codebases; identify gaps and add targeted unit, integration,
and end-to-end tests.
Build and complete test plans for multi-agent orchestration, tool-calling workflows, and multimodal (audio/video/image) understanding features.
Evaluate and benchmark Conversational AI models for accuracy, efficiency, and latency using project evaluation frameworks.
Develop AI agents and LLM-powered workflows to automate repetitive QA tasks such as log analysis, regression triage, and test-case generation.
Identify, document, and regress bugs across the full stack — from containerized deployments and NIM microservices to the frontend client.
Collaborate with engineering and release teams to review feature requirements, specifications, and technical build documents.
What we need to see:
B.Tech. or M.Tech. or equivalent degree in CS/CE/IT/ECE/EEE.
2+ years of hands-on testing experience in software/embedded systems.
Strong programming skills in Python; ability to write
📌 Software QA Test Developer (Pune)
🏢 NVIDIA
📍 Pune