Pavithra Kondavathri

• Collaborated with the development teams for Zoom Phone, Contact Center, ZVA, and WFM to troubleshoot and resolve technical issues. Effectively identified root causes and fixed bugs reported by customer-facing teams.
• Actively contributed to creating workarounds and implementing permanent fixes for identified issues by closely working with the Core Development Team.
• Maintained detailed and comprehensive documentation of issue resolutions, root causes, and implemented solutions, fostering knowledge sharing and driving process improvements across the organization.
• Additional Responsibilities/Continuous Improvements:
  Led the development of an automation tool named “Zoom Phone/Zoom Contact Center Client Parser”, designed to efficiently parse Zoom Phone client memory logs. This tool provides a comprehensive analysis of call flows, SIP ladders (including registration SIP ladders), web requests initiated by the client, response data, and Quality of Service (QoS) metrics.

POC - Intelligent Log Controller to Suppress Log Flooding in OCC
Developed a solution to mitigate log flooding in the Online Charging System (OCC) by intelligently suppressing repeated log statements while ensuring that critical logs remained visible. The project involved researching and implementing advanced logging libraries, including Log4j2, Log back, and the Expiring Duplicate Message Filter (EDMF). The optimized solution utilized Log back's Expiring Duplicate Message Filter, which significantly reduced log file sizes while retaining essential log data and priority information. This implementation streamlined debugging processes and enhanced overall system efficiency.

• Transfer Desk Agent Enhancement: Automated the transfer desk system to provide authors, whose articles were rejected, with suggestions for suitable journals where author's could re-submit their work. This was achieved using Natural Language Processing (NLP).
• Well Care: Developed an automated solution to identify the location of images in a PDF and replace them with corresponding text during the conversion of the PDF into an audio (.mp3) format. This process utilized an Object Detection Model (RCNN) for accurate image recognition.
• Technologies Used: NLP, Object Detection Model (RCNN), and Label-Img for image annotation.