Skip to content
View Abhisri436's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Block or report Abhisri436

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Abhisri436/README.md

πŸ‘‹ Hi, I'm Potri Abhisri Barama!

I'm a Computational Data Science major at NJIT with a minor in Drones and Robotics, passionate about building secure and interpretable AI systems. My work spans machine learning research funded by the NSF and applied AI development through Cornell Tech’s Break Through Tech AI Program. I focus on real-world impact in cybersecurity, healthcare, and financial tech, utilizing ML, LLMs, and model evaluation techniques to drive meaningful solutions.


πŸ›  Tech Stack

Languages: Python, Java, C++, JavaScript, MATLAB
ML & DS Tools: NumPy, Pandas, Scikit-learn, TensorFlow, Keras, Hugging Face, OpenCV, Matplotlib, Seaborn
ML Models: Linear/Logistic Regression, Decision Trees, Random Forest, SVM, KNN, Naive Bayes, Clustering, PCA
DL & LLMs: CNNs, RNNs, Transformers (BERT, GPT), Embeddings, RAG
Dev Tools: Git, VS Code, IntelliJ, Colab, Jupyter
Hardware & Design: Arduino, Microcontrollers, CAD (Fusion 360), 3D Printing, Circuit Prototyping


πŸ“Œ Featured Projects

πŸ” ML-Based Cybersecurity (NSF REU)

Trained ML models on 1M+ samples for IoT threat detection (99% accuracy), and developed a real-time RAG-based LLM mitigation system (~94–98% cosine similarity). [Project Details Coming Soon]

🧠 Dementia Care Alert System (Princeton Hackathon)

Built a full-stack web app with DeepFace to recognize familiar faces in real time (~97% accuracy, <1s latency), with integrated safe-zone tracking for caregiver alerts. πŸ”— GitHub Repo

πŸ“ˆ Insurance Premium Prediction

Developed a Random Forest regression model on 2M insurance records with custom preprocessing; achieved ~1.15 RMSE on Kaggle. πŸ”— GitHub Repo

πŸ“Š Chart Interpretability Research

Analyzed 200+ scientific charts to study how visual and textual elements impact interpretability; designed graphical visualizations to uncover patterns in multimodal chart design. πŸ”— NiiV Lab


πŸ’‘ Interests & Leadership

I'm Co-President of NJIT’s Financial Technology Club and Vice President of NJIT Girl Up, where I lead STEM events and advocate for women in tech.
Outside of tech, I enjoy mentoring in robotics outreach, playing volleyball, and exploring how AI and design intersect to improve real-world experiences. I'm passionate about responsible AI, interpretable machine learning, and using data science to improve outcomes in security, health, and finance.


πŸ“« Let’s Connect!

Thanks for visiting my profile!

Pinned Loading

  1. spam_classification_model_evaluation spam_classification_model_evaluation Public

    Evaluates MLP, SVM, Random Forest, and Decision Tree models on vectorized SMS data for spam detection using accuracy, F1, precision, recall, and log loss.

    Jupyter Notebook

  2. insurance_premium_prediction insurance_premium_prediction Public

    Predicting insurance premiums using Random Forest on a 2M-record dataset with custom preprocessing and model optimization (Kaggle RMSE ~1.15)

    Jupyter Notebook

  3. facial_recognition_API facial_recognition_API Public

    Real-time facial recognition API using DeepFace and OpenCV to identify familiar vs. unfamiliar faces for a dementia care application built at the Princeton Hackathon.

    Python 1

  4. AnanyaTyagi08/Credit-Predictor AnanyaTyagi08/Credit-Predictor Public

    This project implements and compares Logistic Regression, Random Forests, Support Vector Machines (SVM), and K-Nearest Neighbors (kNN) models to determine the best credit prediction model. It utili…

    Jupyter Notebook 1