Software Engineer

ARCHIT AGRAWAL, MSCS

623-***-**** ****************@*****.*** linkedin.com/in/agrawal-archit github.com/architagrawal EDUCATION

MS, Computer Science Aug ‘23 - May ‘25

Arizona State University (ASU) (GPA: 4.0/4)

Coursework: Cloud Computing, Data Intensive System for ML, Data Mining, Principles of Programming Languages, Data Processing at Scale, Mobile Computing, Knowledge Reasoning for AI. BTech, in Information and Communication Technology Aug ‘18 - May ‘22 Dhirubhai Ambani Institute of Information and Communication Technology Coursework: Database Management Systems, Data Structures and Algorithms, Systems Software, Computer Networks, Digital Image Processing, Project Management

TECHNICAL SKILLS

Programming Languages: Python, R, Java, C#, Go, C, SQL, Typescript, JavaScript Frameworks/Libraries: Langchain, spark, Tableau, PowerBI, Llama Index, Gradio, NumPy, Pandas,, Node.js, .NET, Nginx, Redis, MessageQueue, Matplotlib, TensorFlow, scikit-learn, Seaborn, Flask, django, React.Js, Express.js, Ruby on Rails. Databases/Tools: MySQL, AWS EC2, SQS, S3, and lambda, VSCode, Git, Excel, Postman, Docker, Hugging Face, Prompt Engineering, OpenAI API, Linux, Kubernetes, Ansible, Nginx, Redis, RabbitMQ. PROJECT

Reverse-Mode Automatic Differentiation Python, CUDA Feb ‘24 – March ‘24

● Implemented reverse-mode auto-differentiation to adjust weights in models using key concepts and data structures like computation graph and Node, operators such as Add, MatMul, Ones and construction of gradient nodes given forward graph.

● Added kernels for cuda GPU graph executor that can train simple neural nets such as multilayer perceptron models. Handwritten Digit Recognition using TensorFlow and MNIST dataset Python Jan ‘24 – Feb ‘24

● Programmed CNN model to improve on logistic regression model to recognize handwritten digits achieving 99.2% accuracy employing 4 hidden layers.

Soccer Game Result Prediction Python, Deep Learning, Data Science, statistics Oct ‘23 – Dec ‘23

● Increased soccer game’s result prediction accuracy by 12% using LSTM, RNN, Random Forest classifier with XG Boost to supplement past game’s result with user tweets sentiment analysis and game’s bet data. Image Recognition as a Service AWS S3, AWS EC2, AWS SQS, Python Jan ‘24 – March ‘24

● Developed an elastic cloud application using AWS EC2 and AWS SQS for automatic linear scaling based on demand while limiting the maximum number of instances to 20 and queuing all pending requests once the limit is reached, providing image recognition service through deep learning models, serving 100 concurrent requests in 5 seconds.

● Implemented an image recognition service that can take image files as input via a REST API, process them through the deep learning model, and provide the top 1 recognition result in plain text to the users. PROFESSIONAL EXPERIENCE

Student AI Software Developer, Arizona State University Sept ‘23 – Present

● Leading development of corrective retrieval-augmented generation(RAG) based chatbot system addressing hallucinations, facilitating 1000+ faculties to design courses for 60,000+ students. (Python, Langchain, OpenAI, Vector Database, AWS)

● Analyzed student course evaluation and inferred possible improvements in course. (OpenAI API, chatgpt, pandas)

● Designed and implemented APIs and webpages for quiz platform and question bank in ASU Online courses. (Python)

● Supported the functionality and performance of backend services, contributing to the smooth operation of ASU online courses. Software Engineer, Zeus Learning Jan ‘22 – July ‘23

● Refactored the monolithic backend services into various microservices following SOLID principles to optimize and scale the infrastructure using Kubernetes, resulting in a 20% reduction in resource usage and a 35% cost reduction. (.NET, C#, MessageQueue, Redis, AWS S3, Nginx, Docker, Kubernetes, RabbitMQ)

● Optimized SQL queries for fetching data in the class-details web page, resulting in 30% faster screen loading and 10% improvement in the average latency. (Angular, JavaScript, AWS S3)

● Developed a data pipeline and ML prediction system to optimize space reservation systems for booking desks and meeting rooms for 300+ locations, enhancing occupancy rates and facilitating efficient booking amidst COVID-19 restrictions for Goldman Sachs and Merck. (Python, LSTM, Random Forest Classifier)

● Formulated and integrated a custom node package for retrieving N-latest messages from Slack channels, including attached media, documents, and reactions into an internal social networking web app for employees. (Node.js, npm) Software Developer Intern, EAT.FIT Sept ‘21 – Dec ‘21

● Implemented a human-centered order tracking system for the cloud-kitchen-based e-commerce company, optimizing driver location tracking to reduce costs by 24% and enhance customer experience. (React.js, Google Maps API, Python)

● Automated system to scrap competitor product details and reviews using script to better understand business data analytics, positioning of products, and better customer service following Software development life cycle. (Python)

Contact this candidate