• Architecting and developing cloud-native applications using an extensive AWS services stack:

• Event processing with Lambda, SNS, and SQS for real-time notification delivery

• Data persistence with DynamoDB and appropriate S3 bucket tiers for efficient storage

• Infrastructure as Code using CloudFormation for repeatable deployments

• Monitoring and observability through CloudWatch metrics and alarms

• Leading DevOps and API management for LLM-powered chatbot infrastructure:

• Prototyping a “touchless” LLM-powered web application using AI-assisted development and MCP servers to reason over live warehouse health data, emphasizing practical, production-oriented AI use cases

• Developing and maintaining (and prompt engineering) for Python Lambda functions for LLM inference

• Evaluating various foundation models through Amazon Bedrock and SageMaker

• Building APIs to handle model interaction, response generation, and context management

• Implementing monitoring and validation for model performance and response quality

• Designing cross-account and cross-team monitoring solutions enabling metric and log discoverability across Amazon Robotics services, improving operational visibility and reducing MTTR

• Building scalable event-driven systems processing 1000s of events daily for ticket management and notification delivery across AR teams and services

• Developing RESTful APIs through API Gateway to expose critical operational data and enable seamless integration between internal services

• Implementing comprehensive monitoring solutions with custom metrics, automated alerting, and cross-region data aggregation

Key strengths:

• Cloud architecture and AWS service implementation

• LLM integration and API development

• DevOps and infrastructure automation

• Cross-functional technical leadership

• Real-time system design and optimization

• Data pipeline development

• Infrastructure as Code

• Led development of algorithmic solutions to optimize package sortation throughput at Amazon's largest sortation facility

• Designed and implemented simulation-validated algorithms to optimize robotic and manual station utilization, incorporating real-time feedback loops and dynamic load balancing to maximize operational efficiency

• Developed React-based tools for real-time system parameter adjustment and performance testing, enabling operations teams to fine-tune sortation algorithms without engineering intervention

• Created comprehensive testing framework using Python to validate algorithm performance across various operational scenarios and load patterns

• Built data pipelines to collect and analyze performance metrics, providing actionable insights for continuous system optimization

• Collaborated across multiple Amazon organizations including Amazon Air, Operations, and Engineering teams to drive technical decisions and align on implementation strategies, leading to successful deployment across multiple facilities

Developed a data analytics prototype and customer-facing website that provided recruiters and hiring managers with actionable insights on candidate engagement with brand content, supporting data-driven hiring, marketing, and recruitment strategies.

Built several dynamic, real-time pages to present key engagement metrics, including click rates segmented by user demographics, content performance benchmarking, and employer brand comparison by industry and company size, helping clients optimize brand messaging and attract high-fit candidates.

Designed features for evaluating campaign effectiveness across multiple dimensions—content type, topic, author, and recruiting channel—enabling users to refine targeting and track recruitment marketing success across platforms in one unified view.

Leveraged a full-stack tech stack (Python, Django, PostgreSQL, JavaScript, D3.js) to filter and visualize data interactively, eliminating the need for manual data aggregation and analysis, reducing client reliance on spreadsheets, and providing quick and intuitive data insights.

Enhanced the centralized analytics platform that supports recruitment marketing by giving users real-time visibility into candidate attraction, engagement, and brand alignment, empowering data-driven decisions and more effective recruitment strategies.

Teaching Fellow for Introduction to Computer Science (COMP11).