Senior Ai/Ml Developer
CurrentSpearheaded the integration of Nvidia's Multi-Instance GPU (MIG) technology into AI infrastructure frameworks, enabling fine-grained GPU resource allocation and isolation for multi-tenant and multi-workload environments, and by leveraging MIG's capabilities for GPU partitioning and virtualization, optimized resource utilization and improved scalability for AI workloads on Nvidia's GPU-accelerated platforms, unlocking new levels of efficiency and flexibility.Integrated TensorRT with NVIDIA NeMo to optimize text-to-speech models for real-time voice generation in NVIDIA’s AI virtual assistants. Enhanced voice quality and reduced inference latency by 25%, providing seamless voice interactions across multiple languages.Optimized deep learning models for NLP using NVIDIA NeMo and TensorRT, accelerating inference performance for language translation and summarization tasks. Reduced inference times by 35% while maintaining high translation accuracy across multiple languages.Leveraged NLTK and OpenNLP to develop custom NLP pipelines for text analysis, sentiment analysis, and named entity recognition (NER) tasks. Designed and implemented NLP algorithms and models to extract insights from unstructured text data, enabling advanced text mining and information retrieval capabilities.
