Speech LLM Engineer, Voice-First Agentic AI

NVIDIA's Speech team seeks a Senior Speech LLM Engineer to develop spoken language and multimodal AI systems that power secure, scalable, and sovereign voice-first AI solutions.

The role involves:

• Developing and advancing speech, audio, and multimodal LLMs for next-generation voice-first agentic AI experiences.
• Leveraging AI-native and agentic workflows to accelerate research, development, evaluation, and deployment of AI systems.
• Designing and deploying AI models and platforms with strong consideration for AI safety, security, privacy, and Sovereign AI requirements.
• Establishing and driving benchmarking frameworks for voice agents.
• Leading technical initiatives, mentoring engineers, and fostering a One Team culture.

Requirements:

• Master's degree in Computer Science, AI, Electrical Engineering, or related field (or equivalent experience).
• 2+ years of experience building and deploying speech, multimodal AI, or LLM systems.
• Strong Python skills and experience with PyTorch or TensorFlow.
• Hands-on experience with ASR, TTS, speech understanding, audio-language models, or multimodal LLMs.
• Experience building production ML pipelines and MLOps infrastructure.
• Proven technical leadership and mentoring experience.

Preferred skills:

• Hands-on experience with NVIDIA AI technologies such as NeMo, NeMo Agent Toolkit, Nemotron, Riva, NIM, and Voice Chat.
• Experience building voice-first agentic AI systems with reasoning, tool use, and multimodal capabilities.
• Strong expertise in speech AI, including ASR, TTS, speech-to-speech, and conversational AI.
• Experience benchmarking AI agents for quality, latency, reliability, safety, and user experience.
• Familiarity with Sovereign AI, enterprise AI deployment, and data governance requirements.

NVIDIA offers highly competitive salaries and a comprehensive benefits package.