Jan 2023 – Present · 3 yrs 2 mos · Bengaluru, Karnataka, India

• Developed a Spring Boot-based Microservice Template Library, streamlining microservice architecture for VMware on Cloud
• Reduced engineering hours by 34% by creating a one-stop, feature-rich solution for microservice creation using preset Docker base images, standardized Dockerfiles, and templated Helm charts, integrating it with distributed tracing and metrics collection
• Enabled rapid distribution of changes across 200+ microservices by developing a seamless version upgrade mechanism for the library, reducing upgrade effort by 55% and cutting deployment time from several hours to under 30 minutes
• Reduced triaging time by 27% by designing and implementing a distributed tracing architecture, accelerating microservice debugging and upgrade cycles
• Implemented a highly scalable Slack and Email-based notification system with automated Jira creation, reducing manual failure triaging effort and ensuring precise delegation to the correct microservice team
• Contributed to the REST API development of multiple microservices, improving scalability by 3x while maintaining minimal latency
• Enhanced an in-house Concourse-based CI/CD system to integrate automation, patch, and vulnerability fix releases
• Tech Stack: Java, Spring, Springboot, Python, Bashscript, CI/CD, Jenkins, Concourse

Aug 2021 – Jan 2023 · 1 yr 5 mos · Bengaluru, Karnataka, India

Jan 2021 – Jun 2021 · 5 mos · Hannover, Lower Saxony, Germany

• Worked on an extensive study to analyse state-of-the-art Machine Learning on Graphs models
• Joint Learning involves simultaneously learning a noisy graph and a node classifier, leading to more accurate classification
• Defined custom experimental setups and benchmarks to evaluate the performance of various joint learning models

Aug 2020 – Dec 2020 · 4 mos · Bengaluru, Karnataka, India

• Worked on developing a framework feature to automate running Chaos Engineering Experiments for Seamless Resiliency Testing
• Detected 39% of potential faults across application, infrastructure, and network levels by designing a Chaos API layer

May 2020 – Jul 2020 · 2 mos · Bengaluru, Karnataka, India

• A scalable logging and monitoring system for Bixby Voice Assistant
• Designed the architecture for a scalable logging and monitoring system to analyse logs for Bixby Voice Assistant.
• Reduced log retrieval time by 23% by incorporating Autoscaling for 50+ pods & clusters, Automatic Load Balancing
• Created a logging dashboard with metric visualizations using Kibana.
• Technologies/Tools Used: Docker, Kubernetes, Elasticsearch, Fluentd, Kibana, Metricbeat

May 2019 – Aug 2019 · 3 mos · Winnipeg, Canada

• Analyzed minimally 3-connected graphs to formulate a graph with the maximum theoretically possible upper bound for connectivity
• Developed computationally assisted proofs using SageMath and Python to verify the formulated graph
• Co-authored the research paper with Dr. Ortrud Oellermann and published it in a peer reviewed journal

Jan 2019 – Apr 2019 · 3 mos · Pilani

• Creating a CNN-Based Fault Diagnosis Solution for Civil Structures
• Implemented a new data preprocessing method to convert the raw 1D signal data from a set of accelerometers placed in a civil structure into perceivable images.
• Developed a new improved LeNet-5 based CNN model with an accuracy of 98% to detect the damage in the civil structure and classify it into varying damage patterns.

May 2018 – Jul 2018 · 2 mos · Bangalore

• Generating Highlights of Soccer Matches using Machine Learning
• Developed a Computer Vision based product to automate the process of highlight generation in soccer matches for ESPN.
• Created a flexible model which could work on any source video length and generate the highlights of desired length.
• Incorporated various sound, image and video processing techniques including shot detection, object detection and audio classification to create an SVM-based model.
• Displaying Career Highlights In Sports Matches
• Developed a Computer Vision based solution to automate the process of displaying career statistics of players during close up shots in real time.
• Utilised image processing techniques like human detection and facial recognition.
• All code was reviewed, perfected and pushed to production.
• Technologies/Tools Used: Python, OpenCV, TensorFlow, PyAudioAnalysis, Scikit-Learn, Matplotlib, Dlib, NumPy