# Harsh Patel

> Software Engineer at Amazon Web Services (AWS)

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/harshpatel

Working as a Software Engineer in the Bay Area, CA. Currently at Amazon Web Services (AWS).

Technical Skills:
•	Programming and Scripting Languages: Java, Python, JavaScript, Go (Golang), C++, Bash.
•	Web Development Languages and Frameworks: TypeScript, React, Node.js, Angular, HTML, CSS, D3.js, PHP
•	Databases: AWS DynamoDB, MySQL, MongoDB, Graph DB(Neptune), Oracle Database.
•	Tools and Frameworks: AWS, Docker, Git, GraphQL, Hadoop, Kafka, Apache Spark, TensorFlow, Jenkins, Ansible, Selenium, Webpack, ElasticSearch, Zookeeper, Firebase.
•	Gen AI / LLM Technologies: Amazon Bedrock, AI Agents & Workflow Orchestration, Amazon Q, Claude (Anthropic), OpenAI GPT Models, Llama (Meta), Gemini (Google)

Github: https://github.com/patelharsh867
Personal Website: https://patelharsh867.github.io

## Work Experience
### Software Development Engineer II @ Amazon Web Services (AWS)
Jan 2024 – Present | Santa Clara, California, United States
- Designed and delivered the backend of AWS Network Security Director, a newly launched AWS service (2025), enabling high-scale AWS organization-level security scans across 300K+ resources per account using a microservices and cell-based architecture for high resiliency.
- Implemented cell-based architecture for multi-tenant isolation and cell-aware routing for improved fault tolerance.
- Designed real-time organization-level insights aggregation system, achieving sub-second query performance (1.13s for 10K accounts) with 100% accurate data through optimized DynamoDB schema and Index design.
- Implemented rollback-on-failure mechanisms with canary-based composite alarms, achieving zero production incidents during major service deployments across 14 AWS regions.
- Developed response enhancement and preprocessing components for an LLM powered security assistant for AWS Network Security Director, tuning prompts and integrating new APIs to improve response accuracy and formatting.
- Designed and implemented an automated evaluation system with metrics calculation (intent classification, function call accuracy) and reference response generation to systematically measure and improve the Q Extension agent performance
- Designed database scaling strategy using dynamic sharding with random distribution for high-throughput operations.
- Led AWS SDK v1 to v2 migration across 4 authentication services (WebAuthn credential management, token generation, authorization services) with zero-downtime deployment using feature flags, reducing DynamoDB query latency by 15% for a tier-1 Amazon Identity Provider service (IdPrism) that manages and authenticates third-party workforce identities.
- Implemented anti-entropy drift detection mechanisms for identity management, enabling data consistency validation across distributed services with multi-region configuration support.

### Software Development Engineer II @ Amazon
Jan 2021 – Jan 2024 | Sunnyvale, CA
- Designed and implemented distributed log merging system for Alexa's near real-time logging platform using AWS ElastiCache, SQS, and ECS Fargate with 30-second aggregation windows, eliminating log fragmentation across 278+ microservices and processing 50 billion server-side request logs daily.
- Built a high throughput distributed log ingestion system processing 20 million+ messages per second, implementing CloudWatch Logs subscription filters with Lambda functions to forward log events to SQS queues for parallel processing and aggregation.
- Built an internal queue system with dedicated thread pools for ElastiCache write operations, implementing log eviction service with SQS-based message processing and comprehensive metrics tracking for high-volume log ingestion.
- Implemented an ElastiCache integration with Redis for high-throughput log caching, implementing compression algorithms, multimap operations with connection pooling, and configurable health check intervals to support near real-time data availability (~2 minutes).
- Optimized auto-scaling policies and ECS cluster configurations for peak traffic readiness, implementing isolated thread pools for operation segregation and configurable TTL for distributed cache management.
- Architected an automated test account provisioning system for Alexa's load testing platform using AWS Step Functions and Lambda, eliminating manual account creation through dynamic resource allocation based on TPS requirements.
- Created serverless workflow orchestration using Step Functions and Lambda for batch account provisioning, implementing Event Bridge callbacks, task tokens for long-running operations, and automated ticketing for large-scale requests (25K+ accounts).
- Led comprehensive API migration for Alexa Test Accounts and Devices Service (rename, deregister, register operations), including role-based authorization framework with upstream API integration.

### Software Engineer @ Autodesk
Jan 2020 – Jan 2021 | San Francisco Bay Area
• Developing a graph-based data platform to provide vulnerability and application metadata data for all products and services at Autodesk that can be queried to obtain data to identify potential security exposures and anomalies using Machine Learning.
• Developed (Backend-Frontend) a Security Dashboard to display vulnerability data and security posture of Autodesk products.
• Built REST APIs to integrate infrastructure and application vulnerability scan data from different scanning tools to provide
stakeholders real-time data for issues reported in the products from vulnerability scanning tools.
• Built a scalable data pipeline to process, store and publish real time vulnerability data for Autodesk products.
• Optimized data ingestion and import from infrastructure scans by 37% leveraging multithreading and multiprocessing.

### Software Engineer Intern @ Akamai Technologies
Jan 2019 – Jan 2019 | Cambridge, Massachusetts
•Developed a search tool using various NLP embeddings and Text Classification techniques on the indexed text data. 
•Built a chat interface to consider user-based feedback and provide additional choices to provide user-specific search results.

### Software Engineer (Full Stack Developer) @ Tata Consultancy Services
Jan 2016 – Jan 2018 | Pune Area, India
•Developed a Content Management and Display web application, where stakeholders can upload videos, text descriptions, and images, the content was pushed to a mobile app that was used to demo various products of the business.
•Created hybrid cross-platform mobile applications for Android and IOS to display revenue stats and KPIs of a company.
•Developed a form builder web application to generate a form based on custom labels and types provided by a user.

### Technology Analyst Intern @ Morgan Stanley
Jan 2015 – Jan 2015 | Bengaluru Area, India
•Developed a wealth management portal using Ext JS and Sencha Touch and displayed hedge fund and investment analytics by implementing different graph visualizations using D3.js.


## Education
### Master's degree in Computer Science
North Carolina State University
Jan 2018 – Jan 2019

### Bachelor's Degree in Computer Science and Engineering
Nirma University
Jan 2012 – Jan 2016

### High School/Secondary Diplomas and Certificates
A.G. High School
Jan 2005 – Jan 2012


## Contact & Social
- LinkedIn: https://linkedin.com/in/harsh-patel-004

---
Source: https://flows.cv/harshpatel
JSON Resume: https://flows.cv/harshpatel/resume.json
Last updated: 2026-03-22