- Built infrastructure for large language models (LLMs) using TensorFlow and PyTorch, focusing on - prompt engineering and fine-tuning to improve machine understanding for an AI startup.- Deployed a cloud-based ETL pipeline on AWS, integrating data from multiple sources to improve data - quality, integrity, and accessibility.- Transformed Electronic Media Records for data warehousing, analytics, reporting, and alerts.- Led the integration of AI and ML services with TensorFlow and Scikit Learn, improving data workflows - and machine learning processes across global teams.- Acted as a liaison between non-technical staff and developers to ensure smooth communication and - successful project delivery.- Researched and explored advanced AI technologies like GPT-4 and BERT, dedicating 8 hours weekly to - staying updated with industry trends.- Oversaw the public release of an app, using MLFlow for managing models, leading to a 25% increase in - revenue and user engagement.- Conducted deep learning experiments on GPUs for real-time image segmentation and object detection in - surveillance videos using TensorFlow and PyTorch- Developed real-time event extraction platforms with Apache Spark and Kafka, enhancing data accessibility and improving operational efficiency.

- Utilized data visualization tools like Power BI and Jupyter Notebook to aid decision-making, drawing - insights from Python-based analytics.- Wrote complex logic sequences to improve AI agent functionality, integrating third-party navigation - solutions with effective Python scripting.- Ensured high application performance and optimized deployment strategies on the Google Cloud Platform - using Python and Flask.- Worked with cross-functional teams to develop software solutions that met business goals and user - needs, leveraging Flask and Python.- Applied NLP techniques to create a real-time event extraction platform, enabling efficient and - scalable data operations.- Helped develop change detection algorithms using advanced imaging and machine learning techniques - with TensorFlow.- Created data strategies and contributed to scaling Kafka infrastructure to handle high-volume - real-time data streams, utilizing Apache Kafka and Spark.- Led the implementation of microservice architectures using Kubernetes and Docker, enhancing system - scalability and resilience.- Mentored junior engineers and data scientists, fostering a collaborative and innovative team - environment.- Decreased database load by 30% by refactoring, partitioning, indexing, and denormalizing data - structures and implementing an update on event cache.- Automated the manual data reconciliation process between different sources, increasing data accuracy and efficiency by 90%.

- Built large-scale data pipelines for client-side logging at Netflix using Spark and Kafka, supporting - personalization and experimentation efforts.- Developed and optimized storage strategies, reducing data footprint and costs by 40% with Hadoop and - Cassandra.- Designed and implemented a near-real-time data ingestion platform for LifeLock using Python and - Airflow, saving millions annually through proactive member notifications.- Collaborated on a SEP-based data platform to enhance predictive analytics for identity theft - prevention using Python and Spark.- Built and managed a real-time data extraction platform handling 7 million events per second, - optimized for SQL queries on streams, using Kafka and Spark Streaming.- Implemented best practices for data architecture and IT security in hybrid cloud environments using - AWS and Docker, ensuring scalable and secure solutions.- Led the initiative to decompose monolithic data movement products into reusable microservices with - AWS serverless technologies, reducing live migration time by 90% and significantly enhancing - operational efficiency.- Developed an ETL pipeline for loading several terabytes of historical Real Estate semi-structured and unstructured data. Integrated with Apache Kafka and SQS to process streaming data for incremental loads.

- Developed an ETL pipeline for loading several terabytes of historical Real Estate semi-structured and unstructured data. Integrated with Apache Kafka and SQS to process streaming data for incremental loads.- Deployed a cloud-based ETL pipeline on AWS, integrating data from multiple sources to improve data quality, integrity, and accessibility. - Implemented best practices for data architecture and IT security in hybrid cloud environments using AWS and Docker, ensuring scalable and secure solutions.