Part of Flume Efficiency team.

Analyzed and visualized Wi-Fi metrics of Amazon Echo devices using Amazon QuickSight

• Designed and constructed a data processor pipeline and dashboard to analyze Wi-Fi connectivity metrics of Amazon Echo & Alexa devices in real-time.

• Managed by OS device analysts to find out average latency grouped by various device properties, faulty and outlier data, and measure service's performance; eliminating manual work needed to debug on-call issues by ~90%.

• Architected and developed the pipeline applying best practices for Infrastructure as code leveraging AWS CDK (TypeScript) for AWS S3, Kinesis Data Firehose, Lambda, Glue, Athena, QuickSight.

CA for instructor Alexandros Biliris.

• Assisted course for approximately 80 students.

• Posted and graded four homework assignments and two exams.

• Proctored mid-term and final examination.

• Held weekly office hours of two-hour duration to help students with four homework assignment doubts and course material.

• Corresponded and mentored twenty students via email, Courseworks, and Piazza.

CA for instructor John Paisley.

• Attended weekly meetings.

• Answered students' questions regarding course content, quizzes, and projects on the discussion forum.

• Reviewed students' assignments (code) and provide instructional feedback.

• Developed supplementary materials to facilitate the learning process.

• Improved the course by students' feedback and performance data.

Team: Pricing and Promotions

• Ideated, prototyped, tested, and documented the HLD and LLD changes to enable an end-to-end pipeline for delivery executives to assess the item exchanged by the customer; increasing doorstep acceptance rate by ~40%.

• Enhanced an offer format to incentivize customers to purchase products from the same brand again by working on microservices across the stack; led to an increase in 3 million GMV.

• Owned the platform to recommend promotions to sellers on the dashboard by extensively utilizing CRUD APIs for MySQL.

• Architected and developed a conversational chatbot to respond to queries raised by business users of the pricing platform; cutting down the number of on-call tickets by ~20% and improved turn-around time drastically.

• Initiated setting up alerts such as disk usage, ping alerts, 5xx responses for system health checks of various microservices utilized in the team.