Vivek Kumar Pal

Experience

• 

  Infosys

  Aug 2014 - Nov 2018

  Technology Analyst

   Developed regression models to predict bad loans based on customer history Using Tableau to Created interactive charts and graphs to study the impact of introducing the use of ERPM data for Ad serving Developing NLP based application to generate automated mortgage letter for customers based on requirement Working with Amazon RedHat and tableau to create extensive reports representing credit and debit history of bank Working with selenium 3.14 along with JAVA to create a robust, Data driven framework to test the functionality of newly developed GMS system Creating and maintaining Jenkins JOBS for end-to-end CI integration of the developed framework Using GSON and JACKSON libraries for automating and testing API level functionalities Developing test cases for API layer and front-end Self-Service portal using selenium 3.14 and JAVA along with maven and TestNG Debugging failed test cases and providing resolution in real time as per the severity level Creating custom alerts to update users in case of any severe functionality issue Using SSH for validating server-side responses using JAVA automation framework Show less

• 

  Times Internet

  Dec 2018 - Apr 2020

  Data Scientist and SDET

   Cassandra Cluster planning which includes Data sizing estimation, and identify hardware requirements based on the estimated data size and transaction volume Designed and implementation by configuring Topics in new Kafka cluster in QA environment. Managing large data ingestion by using Kafka Designing new Kafka consumers and producers under related topic using JAVA Developed a framework to capture tweets in real time for a keyword from twitter and store it in Elasticsearch clusters using Kafka Worked with Solr and Kibana for segregating, managing and mining data and developing cases to collect data from both sources for running load testing and to be utilized in other performance testing cases Developed regression models using TensorFlow and sklearn libraries to predict the ERPM (Estimated revenue per milli) of the pre served Ads for future serving Creating Clustering model using K means clustering for grouping and identifying similar ads for customer recommendation Optimizing the models frequently for better efficiency and incorporating newer changes as per the requirements Implemented deep learning model using keras Working with matplotlib, pandas, sklearn and NumPy libraries Working extensively with MySQL for extracting data Show less

• 

  EY

  May 2020 - Feb 2022

   Working under the capacity of Assistant Manager in the Data and Analytics team, managing and working with clusters of data from different sources Building efficient data pipelines to extract data from different sources (AWS Blob, SQL, MongoDB, etc.) and visualizing them over PowerBI, Matplotlib, Seaborn after intermediate processing Working as a senior python developer in Health Outcomes Platform to develop and create solutions for data cleaning and processing Creating and automating Alteryx data pipelines Development of efficient machine learning predictive models for Future trends prediction (Time series, Clustering, Regression)  Developed web crawlers using beautiful soup and scrapy to mine and scrape Raw data Developing and managing complex data workflows using Apache airflow Show less

  • Assistant Manager | Data and Analytics

    Oct 2021 - Feb 2022

  • Senior Analyst | Data Science | Machine Learning & Deep Learning

    May 2020 - Oct 2021
• 

  Siemens Energy

  Feb 2022 - Oct 2024

  Team Lead | Machine Learning Engineer

  Working and leading a team of 5+ ML developers |Bearing Metal Temperature: Anomaly detection engine to identify and analyze anomalies within SGT and MGT using bearing temperature and other turbine data>Research on various Tree and Regression based approaches using Python and Scikit-Learn to identify the ideal algorithm using different performance metrices based on the turbine data>Using TensorFlow to design and develop deep learning networks to assist in the identification of the algorithms in the research phase>Building data pipelines using Python to create and test machine data from Snowflake and SE internal Data Lake to be ingested by the developed Machine Learning solution>Designing the AWS based serverless architecture to deploy the developed solution, along with designing suitable REST-API’s using Fast API for integration. Advanced Diagnostic System – Deep: Utilizing Azure OpenAI based large language model to help field engineers with fast and relevant diagnostic solution>Utilizing Azure OpenAI and LangChain framework to create diagnostic chatbot helping field engineers with relevant solutions based on Field service reports>AWS and Azure based solution integration to extract and analyze drawings and texts for reports and use relevant embeddings to improve answer qualityPredictive Emission Monitoring System: Gas turbines Emission Monitoring system to assist site engineers to keep machine emissions within Environmental norms >Worked on Experimenting and analyzing various Machine Learning models using python, Scikit-Learn, TensorFlow based Deep Learning models and documenting results to asses and identify suitable algorithm for prediction output emissions of small, medium and large gas turbines product lines.>Designing and developing suitable data pipelines using Snowflake and AWS S3 to preprocess and ingest data in the Machine Learning system>Developing and deploying on-premise and cloud-based architectures based on AWS Cloud Show less

• 

  American Express

  Oct 2024 - now

  Data Science Manager