Principal Applied Scientist

Microsoft Microsoft · Big Tech · Hyderabad, TS, IN · Applied Sciences

This role focuses on building and leading the development of advanced multilingual speech models, AOAI finetuning, and multimodal generative AI for Microsoft's Azure Speech team. The goal is to create transformative speech technologies for voice agents, transcription, and call centre analytics, impacting billions of users globally, with a special focus on India. The role involves setting technical direction, driving innovation, scaling model quality, and delivering breakthrough technologies.

What you'd actually do

  1. Deliver world-class and transformative speech solutions for Microsoft 1st party and 3rd party products and services.
  2. Set technical directions in multilingual speech model, speech LLMs, model customization, and impact accuracy, latency, and compute.
  3. Build novel data generation solutions to synthesize complex speech scenarios and finetune models.
  4. Build data analysis metrics and solutions to understand the model results, identify gaps, and guide solutions.
  5. Collaborate with the global Microsoft teams, drive innovative solutions for significant customer asks, and deliver sustained large impacts.

Skills

Required

  • BS/MS/PhD Degree in CS/EE or related fields with focus in machine learning, AI, or speech technologies.
  • 10+ years of demonstrated experience in speech or machine learning in academic or industrial setting with skills and aptitude for software design, coding and quality.
  • Demonstration of excellent problem-solving skills in speech and machine learning areas.
  • Proven track record of delivering impactful results and high-quality solutions in complex technical environments.
  • Strong programming skills in Python, C++ or similar languages, with experience in large-scale data processing and distributed computing.
  • Effective communication skills, both verbal and written.

Nice to have

  • Experience with speech/audio processing, multilingual model development, or voice agent technologies.
  • Familiarity with Azure, cloud-based AI platforms, or enterprise-scale deployment of speech solutions.
  • Contributions to open-source projects, patents, or publications in top-tier conferences/journals.
  • Demonstrated leadership in driving technical direction, influencing cross-functional teams, and mentoring peers.

What the JD emphasized

  • advanced multilingual speech models
  • AOAI finetuning
  • multimodal generative AI
  • speech recognition
  • generative AI
  • scale model quality
  • breakthrough technologies
  • AI speech technologies
  • multilingual speech model
  • speech LLMs
  • model customization
  • novel data generation solutions
  • finetune models
  • data analysis metrics and solutions
  • customer asks
  • sustained large impacts
  • academic or industrial setting with skills and aptitude for software design, coding and quality
  • excellent problem-solving skills in speech and machine learning areas
  • Proven track record of delivering impactful results and high-quality solutions in complex technical environments.

Other signals

  • multilingual speech models
  • AOAI finetuning
  • multimodal generative AI
  • speech recognition
  • generative AI
  • scale model quality
  • breakthrough technologies
  • AI speech technologies