SOFTWARE & MACHINE LEARNING

• Built Machine Learning models by researching, implementing, and evaluating a variety of published approaches and algorithms for audio signal processing.

• Played a crucial role in the development and architecture designing of Large Language Model (LLM)-powered chat service (My Assistant) for 2 million+ Walmart employees, managing 10,000+ daily requests with a 99% uptime. Additionally, contributed to the integration of new experiences add-ons for those services.

• Lead the development and deployment of multiple machine learning workflows on Apache Airflow using CI/CD deployment techniques for Git repo, while ensuring integrated logging and monitoring on Splunk, Prometheus, and Grafana. These workflows drove significant improvements in operational efficiency and accuracy.

• Built and fine-tuned multiple Natural Language Processing (NLP) models using state-of-the-art techniques such as TensorFlow and sklearn. These models served as a superior alternative to the existing production model, boosting output accuracy for enhancing customer experience.

• Optimized the performance of existing production workflows, delivering a 533.3% increase in time efficiency and reducing memory utilization by 80% on average. This helped Walmart save significant cloud cost.

• Constructed a comprehensive Knowledge Graph on Neo4j, which effectively organized associate data for faster data retrieval and inferencing. This resulted in more precise and prompt decision-making and analysis, driving operational excellence at Walmart.

• Built Machine Learning models by researching, implementing, and evaluating a variety of published approaches and algorithms for audio signal processing.

• Worked as a Software Engineer in the planning, analyzing, designing, developing, testing, implementing, and maintaining - Software Development Life Cycle (SDLC) for a product.

• Reduced response lag time by 75% by pruning and compressing deep learning models for deployment on edge devices and embedded targets instead of using cloud services, thus increasing model efficiency.

• Recognized anomalies in datasets and boosted model efficiency by 35% by performing data analysis and visualization using PostgreSQL, AWS QuickSight, Grafana, Power BI, and Tableau.

• Built and trained complex deep neural network models with TensorFlow, Keras, PyTorch, and other tools on an AWS EC2 instance. Leveraged data from AWS S3 bucket to train and test the models.

• Developed, integrated, and deployed numerous microservices on AWS Lambda and Docker. Designed and integrated HTTP/RESTful API to access and utilize data, streamlining data processing and enhancing performance.

• Built a Lead Management System using Python to successfully convert incoming leads into new business opportunities and added API integration from Facebook, JustDial, Autoportal, etc. This helped in decreasing lead leakage and increasing annual revenue by 50,000$.

• Created a Business LMS for maintaining business loan customers, appointment, follow up, reporting, and rights.

• Assisted in building an AI chatbot using Python/NLP for the company’s website using Natural Language Toolkit (NLTK) and libraries like scikit, numpy, etc. This helped in increasing the visiting to business lead conversion ratio by 25%.