Role : LLM infra & Optimization- AWS

Location : Chennai, Mumbai, Hyderabad, Pune, Noida, Gurgaon, Bangalore

Role Overview :

Own end-to-end deployment and optimization of large language models on AWS. Drive performance, scalability, and cost efficiency across training and inference workloads.

Key Responsibilities :

- Deploy and scale LLMs using Amazon SageMaker (training, fine-tuning, inference)

- Build and optimize ML pipelines for production workloads

- Improve model performance via GPU/CUDA-level tuning and infra optimization

- Design scalable architectures using AWS (EC2 GPU, S3, VPC, EFS)

- Implement distributed training and multi-GPU orchestration

- Benchmark performance and optimize cost vs latency

- Partner with internal teams/customers on deployment strategy and best practices

Core Requirements :

- 5+ years in ML infrastructure / model deployment / GPU computing

- Strong Python + experience with PyTorch / TensorFlow / JAX

- Hands-on experience deploying or fine-tuning LLMs in production

- Solid understanding of distributed training and inference systems

- Working knowledge of AWS core services (EC2, S3, networking)

Preferred :

- Experience with GPU optimization / CUDA

- Exposure to NVIDIA GPUs (A100/H100) or AWS Trainium/Inferentia

- Knowledge of model optimization (quantization, pruning, distillation)

- Familiarity with MLOps and production ML systems