Biography

Peddamail Jayavardhan Reddy is a Master’s student in Computer Science at the Ohio State University, Columbus. His research interests include Deep Learning, Natural Language Processing and Time-Series Analysis. Prior to his masters, Jay worked at Citi Group for two years, performing the role as an NLP engineer. During the summer, Jay was a Deep Learning intern at The Climate Corporation. Currently, He is a graduate research student under Prof. Huan Sun working on improving code retrieval in software engineering

• Graduate Research Assitant, Nationwide Center for Advanced Customer Insights (Fall 2018 - Present)

  Solving critical problems in Insurance industry using Machine Learning

• Graduate Student Researcher, advised by Dr.Huan Sun (Spring 2018 - Present)

  Currently exploring strategies to use distant supervision techniques to boost the performance of Code Summarization models.

• Deep Learning Intern, The Climate Corporation (May 2018 - July 2018)

  As part of the Sead and Placement team at The Climate Corporation, I worked on the development of a weather-modelling approach for seeding rate prediction.

  • Developed deep unsupervised architecture to generate features from multi-variate time-series weather information.
  • Designed transfer learning approach to use weather features developed, in a supervised machine learning algorithm to predict optimal seeding rate.
  • Developed features from soil composition, hybrids and studied their interaction with weather. The model developed will be deployed as part of the Climate fieldview application.
    

• Natural Language Processing Engineer, Citi Group (July 2016 - June 2017)

  Worked on the developement of Chat Bot for Real-time Financial Trading at Citi group. As part of the NLP team, I performed the following tasks:

  • Developed models for Information Retrieval from financial Trade chats using NLP and Machine learning.
  • Developed regular expressions to capture financial entities from the unstructured financial text data.
  • Developed and tuned classification models to classify between different financial entities (swaps vs bonds).
  • Developed a Named Entity Recognizer to detect entities in a financial chat.
  • Developed a deep learning model which can classify between financial vs social texts and a deep Learning model which can extract Ticker information from an unstructured text.
  • Improved performance of the reference data mapper using Bloomfilters.
    
    

• Application Developer, Citi Group (July 2015 - June 2016)

  Worked on the development Java module for automated Runbook generation.

  • The runbook module reduced the man hours required for runbook creation by 40%.
  • Designed modules for Runbook version control, approval workflow and notification alerts.
  • The runbook module, centralized the data storage providing scope for future data analysis on a large data corpus.
  • Developed the module using core Java, JDBC, Google Web Toolkit(GWT), Hibernate, SQL.
  • Performed POC on blockchain for Reconciliation, proving blockchain helps in avoiding the expensive process of end-of-day internal reconciliations.
  • Presented “Blockchain and it’s use cases in Financial Institutions” at the Citi Annual Town Hall’2016.
    

• Bay Area