Sathwin Reddy
Software Engineer & Cloud Architect
Building scalable cloud infrastructure and full-stack applications. Specialized in AWS, React, and AI/ML systems. Currently pursuing M.S. in Computer Science at ASU.
Professional Journey
My career path building full-stack applications, AI systems, and cloud infrastructure.
Full Stack Software Developer
Full Stack Software Developer
Head of Operations / Technology
- •Leading operations and technology initiatives for ASU's official AI builder community, organizing workshops and hackathons.
- •Coordinating technical projects and fostering collaboration among students building AI-powered applications.
Head of Operations / Technology
- •Leading operations and technology initiatives for ASU's official AI builder community, organizing workshops and hackathons.
- •Coordinating technical projects and fostering collaboration among students building AI-powered applications.
NSF Research Assistant
- •PyTorch multimodal fusion (LiDAR + egocentric camera). CLIP-style contrastive pre-training. 22% zero-shot scene classification improvement on ASU–UCSD benchmark.
- •GPT-4V evaluation harness benchmarking multi-agent perception across 10K+ driving frames. W&B logging turned ablation reviews from qualitative to quantitative.
NSF Research Assistant
- •PyTorch multimodal fusion (LiDAR + egocentric camera). CLIP-style contrastive pre-training. 22% zero-shot scene classification improvement on ASU–UCSD benchmark.
- •GPT-4V evaluation harness benchmarking multi-agent perception across 10K+ driving frames. W&B logging turned ablation reviews from qualitative to quantitative.
Software Engineer Intern
- •Built ML ticket classification service using FastAPI and BERT, improving routing accuracy from 68% to 80%+. Deployed on EC2 with Redis caching and React + TypeScript dashboard for real-time monitoring.
- •Designed automated ETL pipeline combining Python orchestration with C++ components for data-intensive transformations. Integrated S3, RDS PostgreSQL, and Lambda with CloudWatch monitoring for weekly retraining cycles.
- •Containerized analytics application with Docker and deployed to ECS. Built Flask API with Chart.js visualizations to surface model performance metrics.
Software Engineer Intern
- •Built ML ticket classification service using FastAPI and BERT, improving routing accuracy from 68% to 80%+. Deployed on EC2 with Redis caching and React + TypeScript dashboard for real-time monitoring.
- •Designed automated ETL pipeline combining Python orchestration with C++ components for data-intensive transformations. Integrated S3, RDS PostgreSQL, and Lambda with CloudWatch monitoring for weekly retraining cycles.
- •Containerized analytics application with Docker and deployed to ECS. Built Flask API with Chart.js visualizations to surface model performance metrics.
Cloud Developer Intern
- •Built PHI-compliant medical chatbot with React + TypeScript frontend and FastAPI backend on ECS Fargate, connecting to AWS Bedrock for LLM inference with WebSocket streaming for real-time token delivery.
- •Cut document processing costs from $3/page to under $0.01 by rearchitecting PDF accessibility pipeline into serverless Lambda workflow with native C++ parsing library while maintaining WCAG 2.1 compliance.
- •Owned CloudFormation infrastructure-as-code across three AWS accounts and instrumented system with X-Ray distributed tracing end-to-end for rapid debugging.
Cloud Developer Intern
- •Built PHI-compliant medical chatbot with React + TypeScript frontend and FastAPI backend on ECS Fargate, connecting to AWS Bedrock for LLM inference with WebSocket streaming for real-time token delivery.
- •Cut document processing costs from $3/page to under $0.01 by rearchitecting PDF accessibility pipeline into serverless Lambda workflow with native C++ parsing library while maintaining WCAG 2.1 compliance.
- •Owned CloudFormation infrastructure-as-code across three AWS accounts and instrumented system with X-Ray distributed tracing end-to-end for rapid debugging.
Student Representative
- •Served on academic integrity appeal hearing boards, reviewing real cases and contributing to fair, policy-aligned outcomes alongside Academic Integrity Officers.
- •Participated in formal hearings where student sanctions were reviewed, providing student perspective to process with serious academic consequences.
Student Representative
- •Served on academic integrity appeal hearing boards, reviewing real cases and contributing to fair, policy-aligned outcomes alongside Academic Integrity Officers.
- •Participated in formal hearings where student sanctions were reviewed, providing student perspective to process with serious academic consequences.
Teaching Assistant
- •Supporting 120+ students per semester across Python, Java, C++, MATLAB, and data structures courses through weekly lab sessions and office hours.
- •Led lab sessions and ran classes independently when professors unavailable. Provided detailed feedback on assignments and conducted one-on-one mentoring.
- •Recognized with Fulton Outstanding UGTA Award (2023-2024) and GSG Teaching Excellence Award (Spring 2026) for going above and beyond role responsibilities.
Teaching Assistant
- •Supporting 120+ students per semester across Python, Java, C++, MATLAB, and data structures courses through weekly lab sessions and office hours.
- •Led lab sessions and ran classes independently when professors unavailable. Provided detailed feedback on assignments and conducted one-on-one mentoring.
- •Recognized with Fulton Outstanding UGTA Award (2023-2024) and GSG Teaching Excellence Award (Spring 2026) for going above and beyond role responsibilities.
How I build
Full-stack engineering across frontend, backend, cloud infrastructure, and AI/ML systems.
What I've shipped
Full-stack production systems with real infrastructure, measurable impact, and GitHub-hosted source code.
LegalDocs AI
Contract analysis platform built as ASU × Business School capstone. Serverless ingestion pipeline chains Textract OCR into Anthropic LLM via Bedrock. RAG comparison feature lets users ask plain-English questions across contracts.
SmartQueue
Distributed task queue with FastAPI REST API and React real-time dashboard. C++ worker processes for compute-heavy jobs. ECS auto-scaling triggered by CloudWatch queue-depth metrics.
FlowSight
End-to-end real-time traffic intelligence. Kafka decouples bursty camera traffic from the inference API. Grafana caught a checkpoint degrading on nighttime footage that aggregate mAP had hidden.
MediQuery AI
PHI-compliant medical chatbot with real-time token streaming via WebSocket. FastAPI backend on ECS Fargate connected to AWS Bedrock for LLM inference. PDF accessibility pipeline rebuilt serverless with native C++ parser, reducing costs from $3 to under a penny per page.
ChainSwap DEX
Decentralized exchange protocol implementing automated market maker (AMM) with constant product formula. Built TypeScript/React frontend for wallet connection and swap execution. Deployed and tested on Hardhat network with comprehensive unit tests covering edge cases including slippage, price impact, and reentrancy protection.
BookTrack Library
Full-stack library management system with React frontend and Node.js backend. Implemented advanced search and filtering improving retrieval efficiency by 35%. Optimized database schemas reducing query response time by 25%. Integrated JWT authentication for secure user management.
How I run it
Cloud infrastructure, CI/CD automation, observability, and cost optimization across production systems.
Infrastructure as Code
- ·CloudFormation across 3 AWS accounts (AI Cloud Innovation Center)
- ·ECS Fargate + API Gateway + Cognito + KMS per environment
- ·X-Ray trace surfaced exact Lambda fault mid-client demo
CI/CD Pipelines
- ·GitHub Actions → Docker build → ECR push → ECS rolling deploy
- ·Unit tests gated on every merge (SmartQueue)
- ·Multiple deploys/day during final sprint with zero regressions
Observability
- ·CloudWatch alerts caught 2 upstream schema changes before training corruption
- ·Grafana caught checkpoint degradation hidden by aggregate mAP
- ·W&B structured ablation reviews across 3D detection mAP + ADE/FDE
Cost & Performance
- ·PDF pipeline: $3/page → <$0.01 via Lambda + native C++ parser
- ·Query latency: ~2s → <500ms via PostgreSQL index + Redis cache layer
- ·ECS auto-scaling kept SmartQueue latency flat as concurrent load doubled
Awards & Honors
Academic excellence, industry competitions, and teaching recognition.
OpenAI × ASU Pitch Competition — 1st Place
1st place in AI pitch competition co-hosted with OpenAI and Arizona State University
Moeur Award
Top 0.4% of ASU graduates — highest academic distinction awarded at graduation
Teaching Excellence
ASU GSG Teaching Excellence Award (Spring 2026) + Fulton Outstanding UGTA Award (2023–2024). TA supporting 120+ students per semester across Python, Java, C++, MATLAB, and data structures courses.
SUN Award
ASU SUN (Serving University Needs) Award for exemplary service and dedication to the School of Molecular Sciences, recognized for commitment to assisting colleagues during times of need.
Certifications
Cloud and development certifications from AWS, Microsoft Azure, and industry partners.
AWS Solutions Architect Associate
AWS Certified Solutions Architect Associate - designing and deploying scalable systems on AWS
Azure Developer Associate (AZ-204)
Microsoft Certified: Azure Developer Associate - building cloud solutions and services on Azure
AWS Cloud Practitioner
AWS Certified Cloud Practitioner - foundational understanding of AWS Cloud concepts
Full-Stack Developer Certification
ASU/AZNext Full-Stack Developer Certification with Cognizant covering React, Node.js, and modern web development practices
POST /contact
Open to new grad SWE, ML engineering, cloud/infra, and full-stack roles. Graduating May 2026 from Arizona State University with M.S. in Computer Science.
curl -X POST sathwinreddy.dev/contact \
-H "Content-Type: application/json" \
-d '{"role": "SWE | DevOps | ML", "team": "yours", "status": "hiring"}'