7.0 - 12.0 years

0 Lacs

pune, maharashtra

On-site

You will be responsible for leading GenAI projects, overseeing the development and deployment of generative AI solutions from concept to deployment. Additionally, you will need a deep understanding of advanced model architectures like transformers and attention mechanisms. Your role will involve leading the integration of large language models (LLMs) such as Azure OpenAI and OpenAI in cloud environments, particularly Azure services, and vector databases. Furthermore, you will be in charge of ensuring the scalability of AI models to accommodate large user bases across various cloud platforms. Data engineering will be a key aspect of your responsibilities, involving the extraction and preprocessing of data from diverse sources for GenAI processing. Proficiency in integrating AI functionalities into applications, web services, or mobile apps is essential. You should possess knowledge in optimizing model performance and reducing latency for real-time applications, as well as understanding potential vulnerabilities in AI and ethical considerations related to AI deployment. Your role may also require domain-specific knowledge depending on the application, such as in medical imaging or financial forecasting. Skills in continuous integration/continuous deployment (CI/CD) will be necessary for automating the testing and deployment of AI models, ensuring they are always up-to-date and performing optimally. Additionally, you will play a key role in mentoring and guiding junior team members to foster a collaborative and growth-oriented environment.,

Posted 6 days ago

Apply

Senior Distributed Training Research Engineer Krutrim

5.0 - 9.0 years

0 Lacs

karnataka

On-site

As a Senior Distributed Training Research Engineer at Krutrim, you will play a crucial role in training frontier and foundation multimodal large language models. Your primary responsibility will be to develop various generative AI models, including large language models, voice/speech foundation models, and vision and multi-modal foundation models using cutting-edge techniques and frameworks. You will be hands-on in optimizing and implementing state-of-the-art neural architecture, robust training, and inference infrastructure to efficiently bring complex models with hundreds of billions and trillions of parameters to production, focusing on low latency, high throughput, and cost efficiency. Your key responsibilities will include designing and implementing highly scalable distributed training pipelines for LLMs and frontier models, utilizing model parallelism and data parallelism techniques. You will leverage your deep knowledge of CUDA, C++, and low-level optimizations to enhance model training speed and efficiency across diverse hardware configurations. Additionally, you will research and apply cutting-edge parallelism techniques to accelerate model training and reduce computational costs. Your expertise in deep learning frameworks such as PyTorch, TensorFlow, and JAX will be essential for tailoring them for distributed training scenarios and working with massive models to ensure stable and efficient training across distributed resources. You will design and conduct experiments to analyze the impact of model size, data, and computational resources on model performance and collaborate closely with research scientists and engineers to integrate research findings into production-ready training systems. To qualify for this role, you should hold a Ph.D. or Master's degree in Computer Science, Machine Learning, or a related field and have at least 5 years of experience in distributed training of large-scale deep learning models, preferably LLMs or similar models. You are expected to possess a strong theoretical and practical understanding of deep learning algorithms, architectures, and optimization techniques, along with extensive experience in various model and data parallelism techniques. Your expert-level knowledge of PyTorch, TensorFlow, or JAX, coupled with a proven track record of optimizing deep learning models for speed and efficiency using CUDA, C++, and other performance-enhancing tools, will be crucial. Familiarity with current research trends in large model training and the ability to apply new techniques to real-world problems will also be beneficial. Join Krutrim as we shape the future of AI and make a significant impact on hundreds of millions of lives across India and the world. If you are passionate about pushing the boundaries of AI and eager to work with a team at the forefront of innovation, we look forward to hearing from you!,

Posted 1 week ago

Apply

Start Your Job Search Today

Browse through a variety of job opportunities tailored to your skills and preferences. Filter by location, experience, salary, and more to find your perfect fit.