Sophomore at UW–Madison, passionate about building practical and impactful software solutions. I enjoy working across the stack, experimenting with AI/ML, and turning ideas into production-ready applications.
-
In-House Meeting Analysis Platform
- Internship @ Aragen LifeSciences.
- Tech stack: SQL Server Management Studio (database), FastAPI (backend), custom NLP pipelines using spaCy, runtime inference with ollama.
- Designed, developed, and deployed the platform from scratch — successfully conducted the go-live of the web app.
-
Recycling Information Web App (24-hour hackathon project)
- Users can get detailed recycling instructions on any object by uploading a picture.
- Built with Flask, OpenAI API, Firebase, JavaScript, and an object detection pipeline.
-
Gaze & Emotion Detection POC
- Developed at Cognitive Botics.
- Leveraged Haar Cascade classifiers and OpenCV for real-time gaze and facial emotion recognition.
-
IBDP Grade Management System
- Python-based system to manage and track student performance.
-
Email Spam Classifier
- Implemented using Naive Bayes algorithm, trained on the Enron Spam Dataset.
-
Pose Estimation & Classification
- Used TensorFlow’s PoseNet for a workplace inclusivity project.
- Featured at the IB Festival of Hope.
-
Other Projects
- Backend: FastAPI
- Database: SQL Server Management Studio (SSMS)
- NLP: spaCy (custom pipelines), Hybrid Extractive + Abstractive Summarization using TFIDF based extraction with LLM capabilities.
- Inference Engine: ollama
- Deployment: Production-ready web app, API integration
- Backend: Flask
- AI Services: OpenAI API
- Cloud & Auth: Firebase
- Frontend: JavaScript
- Computer Vision: Object Detection
- Computer Vision: OpenCV
- Algorithms: Haar Cascade classifiers
- Application: Real-time emotion & gaze recognition
- Core Language: Python
- System Design: Grade & performance tracking logic
- Machine Learning: Naive Bayes algorithm
- Dataset: Enron Spam Dataset
- Implementation: End-to-end text classification
- Deep Learning: TensorFlow
- Models: PoseNet
- Application: Pose classification for inclusivity project
- LLM Pre training and Post Training techniques
- Efficient and optimized Game development
- GitHub
- 📧 Email: [adityadharap3@gmail.com]
✨ Always eager to learn, collaborate, and create!