Aryan Uppuganti

Aryan Uppuganti

About

Detail

Bengaluru, Karnataka, India
Contact Aryan regarding: 
Flexible work
Starting at USD20/hour
id_card
Internships
Starting at USD1k/month
connect_without_contact
Finding mentors
Finding co-founders

Timeline

work
Job
school
Education
folder
Project

Résumé

Jobs verified_user 0% verified
  • Zeta
    Graduate Intern – Software Test Engineering
    Zeta
    Jan 2026 - Jul 2026 (7 months)
    • Built an AI-powered RAG application using Python, LangChain, and LLMs to generate manual test cases and REST Assured/TestNG scripts from API documentation, reducing manual QA effort by 90%. • Developed Java-based REST API automation frameworks using REST Assured and TestNG, increasing backend regression coverage by 60% while validating payment microservices and Apache Kafka workflows. • Automated 96 end-to-end integration scenarios across rewards, refunds, cancellations, and chargebacks, identifying 198 issues including 30+ functional defects before production release. • Performed production debugging using Kibana, PostgreSQL, and service logs, reducing issue verification time by 40% while improving reliability across distributed payment
  • C
    Machine Learning Intern
    CairoVision Innovations Ltd
    Feb 2023 - Aug 2023 (7 months)
    • Developed Python and OpenCV pipelines for image preprocessing, annotation, feature engineering, and dataset generation, improving dataset quality by 30% for machine learning model training. • Built reproducible computer vision data pipelines using Python, NumPy, and OpenCV to streamline dataset preparation and accelerate machine learning experimentation.
Education verified_user 0% verified
  • B
    B.E. in Artificial Intelligence and Machine Learning
    B.M.S College of Engineering, Bangalore
    Sep 2023 - Jul 2026 (2 years 11 months)
    GPA: 9.13/10
  • Indian Institute of Technology Madras
    Foundational Bachelor’s in Data Science and Applications
    Indian Institute of Technology Madras
    Sep 2023 - Jul 2024 (11 months)
    GPA: 8.00/10
Projects (professional or personal) verified_user 0% verified
  • N
    Network Traffic Analysis & HTTP Load Testing Framework
    Jan 2026 - Current (8 months)
    Developed a Python-based network analysis and stress-testing framework using Scapy, socket programming, and multithreading to inspect TCP/IP, HTTP, and DNS traffic and simulate concurrent workloads. • Implemented packet capture/inspection, latency profiling, throughput analysis, retransmission detection, and automated bottleneckidentification reporting under high concurrent load. • Designed a lightweight monitoring dashboard to visualize latency, throughput, retransmissions, and packet statistics for faster anomaly identification.
  • A
    AI Chatbot Suite
    Jan 2026 - Current (8 months)
    Engineered 6+ AI-powered applications using Python, Streamlit, LangChain, Gemini API, OpenCV, PaddleOCR, and ChromaDB, integrating NLP, computer vision, OCR, and Generative AI into document intelligence, multimodal search, and conversational analytics workflows, reducing manual processing effort by over 80%
  • C
    CardioVision – Cardiovascular Disease Risk Prediction
    CardioVision
    Jan 2026
    • Engineered an end-to-end cardiovascular disease prediction platform using Python, XGBoost, feature engineering, MLflow, Streamlit, and GitHub Actions, achieving 86% prediction accuracy while enabling reproducible experimentation, automated CI workflows, and real-time inference.
  • A
    AI Document Analysis Assistant
    Jan 2026
    • Built a Retrieval-Augmented Generation (RAG) application using FastAPI, LangChain, ChromaDB, PaddleOCR, and OpenAI/Gemini APIs, enabling semantic document retrieval and contextual question answering over enterprise PDF documents through scalable REST APIs.
  • D
    Domain-Adaptive Fine-Tuning of Qwen2.5-3B-Instruct using QLoRA
    Jan 2026
    • Fine-tuned the Qwen2.5-3B-Instruct Large Language Model using PyTorch, Hugging Face Transformers, PEFT, QLoRA, and MLflow, improving domain-specific instruction-following performance through parameter-efficient fine-tuning, experiment tracking, and systematic model evaluation.