ML Researcher / Engineer – Voice & Generative AI

Location: Bengaluru (Work from Office Domlur)

Team: AI & Machine Learning

Experience: 27 Years

About the Role

We are looking for a passionate ML Researcher / Engineer to build and scale next-generation LLM-powered voice and multimodal AI systems.

This role sits at the intersection of Large Language Models, Speech AI (TTS/STT), and Generative AI, powering real-time AI agents used by millions of users.

If you love experimenting with cutting-edge models and shipping them into production at scale this is for you.

What Youll Do

• Fine-tune, optimize, and deploy LLMs, TTS, STT, and voice models for real-time conversational systems.
• Transform messy, real-world audio and text data into structured, high-quality datasets for training and evaluation.
• Build and optimize low-latency inference pipelines for scalable voice agents and multimodal AI systems.
• Collaborate with infra and product teams to ship production-grade GenAI systems with monitoring, observability, and fallback mechanisms.
• Experiment with GANs, diffusion models, audio generation, and multimodal fusion architectures.
• Own the full ML lifecycle research, experimentation, training, evaluation, deployment, and iteration.

What Were Looking For

• 27 years of hands-on experience in AI/ML roles (startup or product-based experience preferred).
• Strong expertise in LLM fine-tuning, instruction tuning, or pretraining methodologies.
• Experience working with TTS/STT systems such as:
• Whisper
• Tacotron
• VITS
• Exposure to multimodal architectures, generative audio systems, GANs, or diffusion models.
• Strong debugging skills ability to identify and fix model failure modes.
• Proficiency in frameworks/tools such as:
• PyTorch
• TensorFlow
• Hugging Face
• ONNX
• NVIDIA Triton Inference Server
• LangChain

Why Join Us?

• Work at the forefront of LLMs, Voice AI, and Generative Models and ship real production systems.
• Build AI agents that impact millions of users.
• Collaborate with a high-caliber AI and product team.
• Freedom to experiment propose architectures, train foundational models, and own innovation.
• Startup agility with enterprise-scale impact.
• Solve multilingual India-first AI challenges across speech, reasoning, and personalization.