May 2022 – Aug 2022 · 3 mos · San Francisco Bay Area

• Strategic Planning & Analytics | Infrastructure Data Center (IDC)
• Built an end-to-end framework to establish Single Source of Truth for 3200 IDC metrics spread over 500+ Tableau dashboards
• Developed UI using ReactJS for business owners to perform CRUD operations on metrics, thereby reducing 91% query time
• Designed Hive Schema and built Dataswarm pipeline with quality checks to process metric data and populate hive model
• Constructed Nodes/Edges Graphical (NEG) Model using YAML & GraphQL to connect multiple entities with 100K+ data points
• Enabled use of data quality check operator in Unified Programming Model as a Data Constraint for 5K+ community
• Implemented algorithm to detect upstream source tables for 3500 Hive Views using data lineage tracking APIs

Aug 2019 – Aug 2021 · 2 yrs

• Big Data Cloud Platform | Global Markets
• Data Consumption & Production Platform
• Developed metadata-driven pipelines to deliver advanced market research data products based on ‘Fortune 500’ data sources
• Implemented dynamic job orchestration, exploration and processing of 10 million rows of daily data using Airflow, PySpark, Azure Databricks & SQL
• Created Pyspark Sourcing Notebooks for data loading and source-to-target mapping according to agreed data formats
• Built processing, validation and profiling notebooks for data transformation according to various quality and threshold rules
• Data Ingestion Tool
• Developed tool for importing tables from RDBMS (Oracle, MYSQL, Sybase) to the Hadoop Cluster using Apache Oozie and Flask Framework
• Eliminated the need for data scientists to reach out to developers thereby reducing 85% of data onboarding time
• As the application owner, ensured the smooth running of 50 scheduled Oozie Jobs at a daily interval and provide end-user assistance

Jul 2018 – Jul 2019 · 1 yr

• Reporting & Analytics | Risk & Finance IT
• Objective Rating Model
• Tested supervised ML models using SVM and Random Forest Classifier for analyzing & rating employees’ annual objectives
• Achieved an accuracy of 88% on test dataset of 1000 employees across 3 global locations
• Used TF-IDF with n-grams as terms and cosine similarity for finding the important keywords related to each output class & similarity scores
• Developed User Interface for displaying the model and probability prediction of each output class using Flask, HTML, JavaScript and CSS
• News Analytics Model
• Worked on News Analytics Tool aimed to reduce bank losses by focusing on news analysis using machine learning.
• Developed supervised news classification model based on the news headlines using LSTM model of Keras library
• Achieved an accuracy of 82% on 2000 news headlines test dataset for news classification model
• Developed an automatic Acronym-Expansion Matching algorithm for news articles using Python and Regex
• Assisted in creation of Kibana dashboard for news summaries with sentiment analysis and risk indication based on market data patterns

May 2017 – Jul 2017 · 2 mos · Pune/Pimpri-Chinchwad Area

• Risk & Finance IT
• Developed bot for automatic solution replies to issues found during User Acceptance Testing runs using machine learning and data mining.
• Helped build data extraction model for extracting mail contents from Outlook using Python.
• Helped build data transformation model for text processing on extracted HTML data using NLTK libraries and Python.