• Designed an Alert Grouping Algorithm in Python, leveraging AWS services for scalable deployment, to cluster alerts based on log similarity, reducing resource utilization by 25% and improving the efficiency of alerts by 20%.• Developed an end-to-end Machine Learning algorithm for real-time Anomaly Detection in log data, utilizing the Prophet algorithm integrated with Databricks. This led to a 20% reduction in operational costs, a 30% improvement in detecting low-key anomalies.• Built… Show more • Designed an Alert Grouping Algorithm in Python, leveraging AWS services for scalable deployment, to cluster alerts based on log similarity, reducing resource utilization by 25% and improving the efficiency of alerts by 20%.• Developed an end-to-end Machine Learning algorithm for real-time Anomaly Detection in log data, utilizing the Prophet algorithm integrated with Databricks. This led to a 20% reduction in operational costs, a 30% improvement in detecting low-key anomalies.• Built and automated ETL pipelines in Databricks to streamline failure analysis, providing actionable insights and generating periodic recovery statistics. This solution reduced on-call engineer bandwidth by 40 days in 2023, significantly lowering operational overhead.• Designed and deployed a web application for monitoring cloud resource utilization. Utilized the React framework, HTML, CSS, and JavaScript to develop a dashboard interfacing with the MongoDB database. This tool is used daily by test engineers and managers to locate unused resources and generate device health and utilization reports, resulting in at least a 60% increase in device utilization.• Worked under Data Engineering (Log Insights and Analytics) team based on Snowflake, AWS and Data Bricks (Elysium Analytics).• Working on Large Language Models (LLM) utilizing advanced models like Llama 2 and mpt7, empowering users to create natural-sounding queries for interactions with cloud data. Show less

• Engineered an end-to-end NLP model to summarize corporate email messages, reducing costs by 35% and improving model accuracy by 12%.• Built and maintained robust data pipelines in Snowflake, utilizing Python and SQL, and integrated with AWS to optimize data processing workflows. Collaborated closely with the Data Engineering team, leading to a 30% improvement in data processing efficiency.• Researched on semi-supervised ML algorithms for Policy Evaluation, ensuring client contracts… Show more • Engineered an end-to-end NLP model to summarize corporate email messages, reducing costs by 35% and improving model accuracy by 12%.• Built and maintained robust data pipelines in Snowflake, utilizing Python and SQL, and integrated with AWS to optimize data processing workflows. Collaborated closely with the Data Engineering team, leading to a 30% improvement in data processing efficiency.• Researched on semi-supervised ML algorithms for Policy Evaluation, ensuring client contracts aligned with company policies. Show less