🎓 Graduate student at the University of Maryland, pursuing an MS in Data Science (GPA: 3.95).
💻 Passionate about AI/ML, NLP, and recommendation system.
🧩 Experienced in building real-world AI solutions, from LLM-powered apps to large-scale data pipelines.
🏅 Certified in Microsoft Azure AI Fundamentals.
🌍 International student from India 🇮🇳

Languages: Python 🐍 · SQL · C · MATLAB
Libraries/Frameworks: scikit-learn · NumPy · Pandas · Matplotlib · Seaborn · TensorFlow · Hugging Face Transformers · Flask · Gradio
Platforms/Tools: Databricks · PySpark · AWS (EC2, S3, EMR) · Azure ☁️ · Git/GitHub · GitLab · Cursor IDE · Jupyter · Postman · Hugging Face Spaces
Techniques: Machine Learning · NLP · Recommender Systems · Prompt Engineering · Reinforcement Learning · Statistical Modeling · Data Wrangling · Model Serving & App Deployment
New from latest project: CLIP · Cosine Similarity–based scoring · Space YAML config · Lightweight UX for model evaluation

Interactive Gradio app that:
• Computes image–text alignment via CLIP cosine similarity
• Estimates an aesthetic score using reference HQ images
• Deployed on Hugging Face Spaces with a simple, shareable UI

A lightweight Python search engine built from scratch.
Uses TF-IDF ranking, subword BPE, and Levenshtein-based autocorrect to retrieve relevant passages from large text corpora.
Demonstrates practical information retrieval and NLP fundamentals.

• Built a personalized rewards recommendation system in Databricks using Python and PySpark to target merchants based on customer spending.
• Implemented collaborative filtering, explored two-tower deep learning, and tested LLM-based approaches (LLaMA, Gemma, GPT-OSS) for explainable recommendations.
• Delivered insights that helped the AI Team secure buy-in for deployment and future iterations.

• Developed a Spark-based pipeline on AWS EMR to ingest, process, and analyze real-time Bitcoin prices with S3 integration.

• Created a Python-based ML/NLP platform to retrieve research papers, summarize them with BART Transformers, and support automated literature reviews.

• Achieved 99.43% accuracy using Random Forest Classifier with KNN imputation & SMOTE.

• Built an NLP-powered chatbot to interpret 10-K & 10-Q reports for intuitive financial insights.

• Developed a Python + SQL + Flask application to manage pizza orders and streamline operations.

• Advancing skills in LLM applications & prompt engineering.
• Exploring LangChain, RAG pipelines, and model deployment.
• Building hybrid AI architectures combining traditional ML with LLMs.

📧 Email: rithi.basky@gmail.com
💼 LinkedIn

⭐️ Thanks for stopping by! Let’s collaborate and explore the world of AI and data science together.