Business Analyst | SQL • Python • Power BI | MS IT @ ASU | F-1 OPT | Open to Relocate

github

vaishmitha Reddy Biyyapu  • Peerlist

Description: Worked as part of a team to analyze large scale book rating data and build a recommender system using collaborative filtering techniques. The project focused on generating personalized book recommendations based on user similarity and rating patterns. Key Contributions: 1.Cleaned and preprocessed user-book rating datasets using Python (Pandas, NumPy). 2.Reduced dataset to the top active users and books to improve performance. 3.Built a user-based collaborative filtering model using Nearest Neighbors with cosine similarity. 4.Generated top 5 personalized book recommendations for each user. 5.Merged recommendation results with book title data and exported final results to CSV. 6.Applied sparse matrix techniques for efficient big data processing.

Analyzing Big Data Book Recommender System (D

Engineered a cutting-edge solution to enhance road safety by implementing an AI-driven voice-assisted navigation system. Leveraging CNN, a powerful supervised learning algorithm, the project aims to provide real-time guidance in vehicles, reducing the risk of accidents and improving overall navigation.

Obstacle detection

Description: Developed an end-to-end machine learning project to predict Netflix user ratings and viewing preferences using Amazon SageMaker. The project focused on data preparation, model training, deployment, monitoring, and performance evaluation. Key Contributions: Prepared and transformed large-scale user viewing data into LIBSVM format for model training and testing. Engineered key features such as watch time, duration, and watch ratio to improve prediction accuracy. Built both regression and classification models using XGBoost on Amazon SageMaker. Deployed trained models as real-time endpoints for inference. Monitored model performance and system logs using AWS CloudWatch. Evaluated model results using accuracy, precision, recall, and F1-score metrics. Visualized prediction results and user genre preferences for analysis. Managed AWS S3 buckets for training and testing datasets and deleted endpoints after testing to control costs.

Predicting Netflix User Ratings and Viewing T

As part of a group of four students, I contributed to developing a Project Charter for IniTech Solutions to automate their Asset Management Process. My key responsibilities included:Conducting research and collaborating with team members to define the Problem Statement, which highlighted inefficiencies in manual asset tracking and inventory management.Drafting a concise Technology Solution Statement, proposing a scalable and centralized ITAM system to streamline asset management operations.Identifying and documenting quantifiable Project Benefits, such as reducing asset tracking time by 40% and improving inventory accuracy by 25%.Assisting in creating the High-Level Project Plan, ensuring alignment with the Systems Development Life Cycle (SDLC) phases and organizational goals.Preparing a detailed analysis of Technology Tools, recommending the use of Cloud Computing and Relational Databases for enhanced scalability and security. Outcomes Achieved: Developed a comprehensive Project Charter that was well-received for its clarity and strategic alignment with organizational goals.Laid a solid foundation for the ITAM system by providing a clear roadmap for implementation and addressing stakeholder concerns.Demonstrated teamwork and project management skills, which contributed to the successful completion of the project on time.

Project Charter for IT Asset Management Syste

Built a full-stack app that turns raw meeting notes and requirements docs into structured, searchable knowledge  requirements, action items with owners, KPIs, and stakeholders so they stop getting lost after the call ends. - Designed and shipped a FastAPI + SQLite backend with a vanilla JS frontend, end to end - Built a rule-based extraction engine with 6 passing pytest tests, architected behind a clean interface so an LLM-backed extractor (Claude/GPT) can be swapped in with zero changes elsewhere A Business Analyst's take on a problem most companies haven't solved: organizational knowledge evaporates after every meeting unless someone captures it well.