Feb 2022 – Aug 2025 · 3 yrs 6 mos · Bentonville, Arkansas, United States

• SideKick Project:
• Closely working with Data Science and Delivering Location based priorities.
• Pyspark and Dataproc Serverless based deployments to effectively deliver with scale.
• Firefly project:
• Built and delivered Events data ingestion on time to the business and reduced the processing rate from
• 20 msgs/sec to 600 msgs/sec and at a refresh rate between 5-20 sec.
• Migrated the flow to GCP with standardization achieving 7.1k msgs/sec and scalable based on the load.
• Created visualizations to display the freshness of data and implemented data quality standards.
• Implemented dashboard with hourly refresh and keeping it available to the DC level staff to track POs.
• Quantum project: Migrated pipelines from Cloudbreak Cloudera platform to DPaaS, moving from persistent to ephemeral cluster with a huge cloud cost advantage.
• Migrated to new Kafka cluster and monitored using lenses

Jun 2021 – Aug 2021 · 2 mos · Dallas, Texas, United States

• Built end-to-end streaming pipelines for 2 tables with Kafka, bigquery, and monitored topics using lenses.
• Migrated existing pipelines of 11 tables from unmanaged Azure to managed WCNP and brought them to production with git, concord and looper .
• Loaded historical data for 2 tables onto bigquery using sqoop.
• Created visualizations for data completeness monitoring using looker.

Aug 2020 – Jan 2021 · 5 mos · Bengaluru, Karnataka, India

• Extracted and pruned real-world data of microservices health using REST API calls.
• Used weighted average, ARIMA, and exponentially weighted average to predict the failure of microservices on real-world data center setup using Kubernetes and Docker.
• Utilized FBprophet to establish seasonality by day, week, and year in the data and improve predictions.
• Established correlation between various metrics which can reduce the number of metrics to be monitored.

Nov 2018 – Mar 2019 · 4 mos · Tamil Nadu, India

• Extracted features from two ppg waveforms extracted from two proximal sites on the carotid artery.
• Modeled a feed-forward neural network using PyTorch to predict the SBP and DBP using extracted features.
• Experimented with various loss functions and regression techniques and optimized the model.
• Established the accuracy with an RMSE of 4mmHg and 6mmHg for DBP and SBP respectively.

May 2018 – Jul 2018 · 2 mos · Greater Bengaluru Area

• Analyzed device log data using SVM with RBF kernel and Hidden Markov Models for false alarm prediction
• Developed an RNN-LSTM model to handle the sequential nature of data and achieved 78% accuracy.
• Analyzed RCNN, Fast-RCNN, Faster-RCNN, and Yolo to understand the existing bounding box algorithms.
• Designed head detection models using modified YOLO v3 and got 0.70 and 0.74 AP on popular datasets.