● Lead in the evaluation and design of highly complex and innovative software architecture● Creating pipelines, data flows and complex data transformations and manipulations using PySpark with DataBricks● Worked closely with development team members to identify and propose solutions to improve system performance and scalability● Led initiatives to integrate Large Language Models(LLMs) to streamline workflows and boost productivity for development teams

Work collaboratively with senior management to develop strategy an approach to defining business challenges to be answered by business scienceAggressively acquire and train new talent, maintaining a friendly and collaborative work environment , and develop future managers and leadersEstablish effective working relationships with all levels of staff and management Develop strategies to help clients activate their analytics and create a more data-driven cultureProvide thought leadership and emerging quantitative fields where data science can play a significant roleProvide technical mentorship to data scientists and guide technical thinkingHands on implementation of various AI/ML tasks (text clustering(KMeans, KMeans +UMAP, UMAP+HDBSCAN), similarity between titles, session analytics, Extracting intent from short text using BERT and hugging face libraries)Hands on implementation of Support Ticket classificationHands on implementation of Support Ticket Volume forecastingHands on implementation of synonym extraction from dataset, custom text summarization for domain-specific dataset , title generation for long-text, create domain specific taxanomy from short text data, long question answering using bert- Prompt Engineering (blog creation using chatgpt)-Corporate Internal training for AI/ML, group sessions in ML and deep learning-Working on creating functions for Agents using FireCrawl, BitQuery, Livepeer using LlamaIndex/Langchain and no framework- Finetuning computer vision models-Creating pipelines, data flows and complex data transformations and manipulations using PySpark with DataBricks

Managed 30 FTE(Developers) to build open data platform using Databricks and Swaggerhub using AWS which included applying various transfer-learning techniques using pre-trained word embeddings like Glove,BERT for building the search engineEngaged with multiple levels of stakeholders to design, develop and operationalize Key Performance Indicators (KPIs) and Key Risk Indicators (KRIs) which support the tactical and strategic needs of business partners(Department of Homeland Security,Federal Bureau of Investigation and Costal Guard)Managed the team responsible for moving Compliance data into the Cloud (Google Cloud Platform). I pioneered the effort in N16 to move to the Cloud and leveraged it for Reporting & Analytics.Work closely with the Data Office and Technology teams for Project prioritization and Data OversightLead and motivate employees retaining high performing talent with an Enterprise and Diversity lensPosted in Yokota AirBase , Japan as a Data Scientist Specialist

United States Patent and Trademark Office, Alexandria, VA. May 2006-Present I am a senior solution architect for big data and analytics solutions based on Hadoop. For the last 5 years I am working in Hadoop for patent and trademark office, helping and leading the efforts to set up multi node clusters. • Developed data pipeline using Flume, Sqoop, Pig and Java map reduce to ingest into HDFS for analysis. • Hands on experience in installation, configuration, and using Apache, Hortonworks Yarn distributions. • Involved in SQOOP, HDFS Put or Copy from Local to ingest data. • Used Pig to do transformations, event joins, filter bot traffic and some pre-aggregations before storing the data onto HDFS. • Used Hive to analyze the partitioned and bucketed data and compute various metrics for reporting. • Developed Hive DDLs to create, alter and drop Hive TABLES. • Involved in writing Spark jobs using PySpark• Build the regression models with multiple variables, K­means clusters.• Analytic Program Advisor advising Senior Executives on end-to-end program management aspects including vendor selection, package selection, budget & estimates, work plans, staffing, change control, business transformation and training.• Identify and analyze user requirements and recommend appropriate applications or modifications. • Plans, designs, implements and evaluates the work of the functions to address results of objectives against the USPTO and user community objectives, agency priorities, anticipated trends and other similar factors.• Identifies potentially conflicting requirements from different organization and works with senior USPTO managers to resolve conflicts and prioritize requirements.

Learned trouble shooting of various small integrated circuits