Cognitud Advisory

Principal Scientist

Category: Data Analyst & AI

Experience:10–15 years
Location:Pan-India

Your team

You will work with a top-tier organization that's been in the game for 50 years, partnering with some of the world's biggest businesses. As India's largest multinational business group, they boast a workforce of over 500,000 highly skilled consultants spread across 46+ countries. Beyond that, they're at the forefront of the financial markets and data industry, delivering exceptional services in Data & Analytics, Capital Markets, and Post Trade. Their commitment to excellence reverberates in everything they do.

Your job

  • You will lead AI/ML initiatives, focusing on NLP, Generative AI, LLMs, and multimodal models, from research to large-scale deployment.
  • You will drive model training, optimization, and production deployment with enterprise-grade impact.
  • Ability to design and implement scalable AI systems, including distributed architectures, data pipelines, APIs, and integration with vector databases.
  • You should oversee and implement MLOps/LLMOps practices using tools such as MLflow, Weights & Biases, DVC, SageMaker, and Vertex AI.
  • You will mentor and grow high-performing AI/ML teams, fostering innovation and adoption of cutting-edge AI research.
  • Ability to collaborate with stakeholders to translate research insights into commercial AI products and solutions.
  • You will ensure code quality, reproducibility, and best practices in Python frameworks, including Hugging Face Transformers, PyTorch, TensorFlow, Keras, and JAX.

Capabilities

  • You must have 10–15 years of industry/academic experience in AI/ML, with 7+ years in NLP and Generative AI.
  • Experience in large-scale model training, optimization, and deployment of LLMs and multimodal AI models.
  • You should have expertise in Python and frameworks like Hugging Face Transformers, TensorFlow, PyTorch, Keras, JAX.
  • Strong knowledge of LLMs, multimodal AI, embedding models, vector databases, and agentic AI.
  • You should have hands-on experience with MLOps/LLMOps tools (MLflow, Weights & Biases, DVC, SageMaker, Vertex AI) for production-ready ML pipelines.
  • You should have solid system design skills, including distributed systems, data pipelines, and API development for AI solutions.
  • Demonstrated ability to publish top-tier research in LLMs, multimodal models, or related areas.

Qualifications

Bachelor's or Master's degree in Computer Science, Engineering, or a related field.