import aryan_kumar as dev
"""
Computer Science Graduate Student & Software Developer
Passionate about building innovative solutions using Latest Technologies.
Currently pursuing MS in Computer Science at San José State University.
"""
I'm a Computer Science graduate student at San José State University and an SDE Intern at Sun-Net Inc developing User Activity Tracking dashboards. Previously, as an Associate Consultant at Datacurate Technologies, I spent over two years developing enterprise web applications, data marts, and ETL pipelines.
Driven by a passion for backend engineering and AI, my focus is on Java, Python, and LLM integrations. I thrive on transforming complex business requirements into high-performance, maintainable software.
Years Experience
Certifications
Graduate GPA
Developed dashboards for tracking software-related activities consisting of commits and documentation for real-time data tracking.
Built database schemas and design to efficiently store data. Integrated LLM functions used in chatbot to get insights from user activities and generate a comprehensive report with in-depth analysis.
Currently pursuing advanced studies in Computer Science with focus on Machine Learning.
Collaborated within a team to develop an online policy generation platform for an insurance firm using Flask. Contributed to building of key modules and APIs including referral system, payment tagging and premium generation. Optimized system performance by reducing database calls using OOP concepts and restructured schemas to eliminate redundancy.
Developed data marts and management routines using SAS DI Studio, including automated ETL jobs for scheduled DWH refresh, data cleansing, and aggregation processes. Built dashboards and profitability reports for data visualization and monitoring using SAS Viya and EG. Developed customer acquisition websites using the Django framework.
Completed undergraduate studies with a strong foundation in computer science fundamentals and software development.
Completed 12th grade with focus on Science stream including Physics, Chemistry, Mathematics, and Computer Science.
Completed 10th grade with excellent academic performance and strong foundation in core subjects.
MLflow
Hugging Face
Maven
Claude
Gemini
Microsoft
2024
IIRS, ISRO
2022
Udemy
2022
Bolt IoT
2022
Microsoft
2022
Coursera
2021
NPTEL
2021
Coursera
2021
Developed a full-stack job application tracker featuring a Manifest V3 Chrome Extension and Next.js dashboard for one-click job application ingestion for tracking. Integrated Google Gemini Flash API server-side with structured prompt engineering and multi-tier fallback parsing algorithms.
Building a unified platform to streamline the workflow of student group assignments with MERN stack. It will be integrated with real-time chat rooms, task management, and role-based project tracking.
Performed Transfer Learning to fine-tune a pre-trained DistilGPT-2 model by optimizing hyperparameters like learning rate, warm-up steps. Designed a custom Data Collator and training loop to handle variable-length sequences and special tokens for structural guidance to feed the transformer.
Developed a full-stack AI chatbot to provide personalized post-operative guidance and automated triage using patient medical records. Used FastAPI in backend, RedisVL in vector search for RAG, Gemini API in LLM for contextual chat and triage via RAG.
Trained a TensorFlow-based Bidirectional LSTM for the task of natural language generation, specifically song lyrics generation. The model learned complex lyrical patterns from a multi-genre Kaggle dataset of English songs to create original verses.
Built a Python program to convert text into morse code with signals transmitted using Bolt IoT module. Utilized cloud platform for communication between users with real-time messaging capabilities.
Created a machine learning model which catches a credit card defaulter based on input features such as yearly debt payments, credit score, net yearly income, etc. Part of American Express Hackathon. Decision Tree was used with accuracy of 97.93%
Created a machine learning model which predicts the drug required for the patient given the patient's vitals such as BP, Cholesterol level, etc. Logistic Regression was used with accuracy of 93.3%
Designed and implemented a vehicle safety prototype using Arduino Uno, motors, and sensors to enhance accident prevention and passenger safety. Developed the embedded code to calculate distance between the car and surroundings using sensors and automatically apply brakes based on vehicle speed.
I'm always interested in new opportunities and exciting projects. Feel free to reach out if you'd like to collaborate.