Oct 2021 – Jun 2022 · 8 mos · Bengaluru, Karnataka, India

• Developed Bulk Operation feature with complete backend and frontend implementation for admins to manage Global Decision Trees, which provides guidelines and solution steps around resolving customer complaints. Reduced avg time spent by admins to maintain GDT by 20%
• Worked on cloud migration of Ingestion Service which captured status changes in customer-agent conversations through push-pull mechanisms utilizing api calls and cron-jobs and logged key details over ELK Stack crucial for business analysis
• Worked on configuring different roles in Cyclops, a one stop destination for all customer agents to view/modify user and order details
• Developed timer indicators in Pizza, a chrome extension which conveniently provides customer agents with all the necessary details to resolve a customer issue along with ability to close/wrap up tickets by seamlessly integrating with a third party ticketing app

Mar 2021 – Oct 2021 · 7 mos

• Designed and developed the solution for transitioning from a request-response pattern to streaming pattern, incorporating multi-threaded timers and asynchronous processing
• Integrated Liquibase for database schema management and version control for easy scalability for increasing number of config requests
• Developed an event handler library to process and manage all configurational data provided by SDC, crucial for processing payment flows.
• Mentored interns for the batch of 2021, devising their problem statement and work plan for the duration of 2 months
• Part of the Payment Networks Mentoring Program, mentoring people on concepts like OOPs, Java, Spring Boot, and architectural designs

Jun 2019 – Mar 2021 · 1 yr 9 mos

• As part of an innovation project, developed a working model which could differentiate between different debit cards using image processing and deep learning algorithms with an accuracy as high as 96%
• As part of an innovation project, built an intruder alert system capable of detecting a person entering a room, incorporating facial recognition algorithms, and using Raspberry Pi and Infrared sensors
• As part of the performance engineering team, did comparative performance evaluation of different business units in Core Payments system
• Developed and deployed Switch Data Connector, an event driven application responsible for delivering consistent reference data to underlying business containers in the Network of Future space as part of the payment Authorization and Clearing flows
• Developed a data chunking library to allow for seamless breaking and accumulating of big data, reducing the overall network load and processing times for data transferred through event channels

May 2018 – Jul 2018 · 2 mos · Ann Arbor, Michigan, United States

• Measuring Central Black Hole Mass of a Galaxy (Guide: Prof. Dr. Monica Valluri)
• Focusing on one elliptical galaxy, utilized Multi Gaussian Expansion technique to model its luminosity density profile
• Incorporated voronoi binned kinematical data by representing the line-of-sight velocity distribution of each spaxel with truncated Gauss-Hermite series, further convoluted with a point spread function
• Successfully performed the Schwarzschild modeling with a library of 8100 orbits and estimated the black hole mass
• The work was crucial for the goal of fully understanding the connection between galaxies and their nuclear black holes

May 2017 – Jul 2017 · 2 mos · Edmonton, Alberta, Canada

• Galex Ultraviolet Counterpart Matching in the Galactic Bulge Survey (Guide: Prof. Dr. Craig Heinke)
• Utilized low UV field densities to constrain optical and IR counterparts for Spectral Energy Distribution (SED) analysis
• Successfully classified the sources into single stellar, non-compact binaries and compact binaries using images and SEDs,
• identifying 25 compact binaries using UV excess
• High resolution spectroscopy of such sources in future, influential in a better understanding of the evolution of the Universe

May 2016 – Jul 2016 · 2 mos · Göttingen, Germany

• Noise Effects in Spectropolarimetric Inversions (Guide: Prof. Dr. Sami K. Solanki)
• Simulated multiple models of Stokes profiles of Sun for regions like umbra and penumbra, implementing ME inversions using Markov Chain Monte Carlo (MCMC) technique to make claims on the strength and orientation of the magnetic fields
• Successfully proposed the idea of fitting multiple lines in the same model as a way to increase the signal to noise ratio and inferred the noise levels in observed data by taking it as a free parameter in the model
• The work was instrumental in the latest proposal for the launch of the balloon-borne solar observatory “Sunrise” by MPS