Roshan Joseph

I'm an ECE undergrad at NIT Trichy who loves turning complex data into usable tools. Whether I'm building end-to-end ML pipelines or experimenting with LLMs, I enjoy the entire process of bringing an idea to life. I'm currently looking for a summer internship in data science or machine learning where I can contribute my skills in Python and SQL to solve real problems.

• Technical Documentation Assistant: Engineered a Retrieval-Augmented Generation (RAG) pipeline using LangChain and FAISS for vector search, integrated with a locally hosted LLM to query technical manuals.
• Operational Risk Prediction System: Built an end-to-end ML pipeline in Python (preprocessing, training, evaluation) and deployed the predictive model as a scalable REST API using FastAPI.
• Visual Question Answering Tool: Implemented multimodal image processing by extracting high-dimensional visual features with CLIP and synthesizing them with an LLM for natural language responses.
• Research Paper Recommendation Tool: Constructed a content recommendation engine by generating semantic NLP embeddings and applying similarity search algorithms to match relevant academic abstracts.
• Business Intelligence Forecasting Dashboard: Developed an interactive data application using Streamlit and Pandas to visualize KPIs, integrating Scikit-learn models to generate real-time business forecasts.