See this at: https://varunsdevang.github.io
- Location: Amherst, MA
- Email: varundevang@gmail.com
- Phone: (413) 404-2490
- LinkedIn: linkedin.com/in/tsvarun
- Degree: Masters in Computer Science
- Duration: Feb 2023 - Dec 2024
- GPA: 4.00
- Relevant Coursework: Distributed Systems, Software Engineering, Reinforcement Learning, Internet Security, Advanced NLP, Robotics
- Degree: B.Tech in Computer Science and Engineering
- Duration: Aug 2016 - Aug 2020
- GPA: 8.31/10
- Relevant Coursework: Data Structures, Computer Networks, Cloud Computing
- Duration: Aug 2020 - Dec 2022
- Developed, designed, and maintained Assess services, an online drone imagery tool for property assessment.
- Implemented scalable Go Rest and GRPC microservices for handling geospatial data, images, and large binary data with authentication and access control.
- Achieved a 50% reduction in latency by implementing GRPC stream APIs for large binary data transfers, conducted load testing, identified critical bugs, and performed fixes.
- Integrated AWS Batch processes, reducing container load by 15% and significantly cutting AWS costs. Acquired proficiency in C# and delivered critical release APIs within a week.
- Designed and modeled an event-based internal notification platform utilizing AWS SNS and SQS, capable of handling over 1000 requests concurrently.
- Pioneered a serverless service orchestrator, replacing outdated technology with AWS Step Functions and Lambda, resulting in a 60% cycle time reduction.
- Translated critical code from MATLAB to Python, increasing scalability and reducing latency by more than 60%.
- Orchestrated cross-functional teams to establish standard interface models and proposed OpenTelemetry integration for serverless architecture tracing.
- Duration: Jan 2020 – July 2020
- Programmed scalable Go microservices, improved API documentation for resource onboarding, and proposed ELK stack integration for enhanced system tracing.
- Initiated a side project to prototype a Deployment Dashboard, leveraging Jenkins, HTML, CSS, JavaScript, and AWS S3 to monitor multi-environment deployments.
- Developed a human-following capability for a holonomic robot equipped with an RGB-D camera using YOLO-v7, SuperPoint, and a PID controller.
- Fine-tuned Transformer models like BERT, GPT-2, and T5 for multi-label classification of quotes. Experimented with various prompting techniques on T5 and Gemma models for the quote interpretation task and analyzed the results.
- Prototyped a simple stock trading application in Python to test distributed systems concepts like multithreading and leader elections.
- Developed and implemented custom machine learning models—Neural Networks, Random Forests, and K-Nearest Neighbors (KNN)—for complex data analysis tasks.
- Leveraged hyper-parameter tuning and optimization techniques to maximize model performance, creating F1 Score graphs for data-driven decisions.
- Programming: Golang, Python, SQL
- Software: AWS, Docker, Kubernetes, Jenkins, Jira, Confluence, Postman, Git, Postgres, DynamoDB
- Concepts: Debugging, Data Structures and Algorithms, OOP, Web Development, System Design, Software Development, Testing, Agile Methodologies, Serverless Design, Microservices, Machine Learning
For further details, you can reach out via email or connect on LinkedIn.