Roshani kumari

I’m a passionate data scientist and AI agent developer with a strong foundation in Machine Learning, NLP, and Deep Learning, focused on solving real-world ML problems with engineering discipline. I’m eager to turn analytics and models into scalable AI solutions.

As an AI Agent Developer Intern at Pulp Strategy, I automated technical SEO audits by building an AI agent that checks robots.txt, XML sitemaps, and crawl errors using Python and web automation (n8n). I also built pipelines for Page Speed Insights, image optimization, and schema validation (FAQ, Product, Breadcrumbs), improving detection and reporting efficiency by 30%.

I’ve also built end-to-end ML projects like a MentalFitnessTracker using supervised learning (Linear Regression and Random Forest) with Pandas, Matplotlib, and scikit-learn, then delivered personalized recommendations via Google Colab. In my projects, I designed a FastAPI-powered AI GitHub PR reviewer using Llama 3.3 via Groq API and created a microlearning recommendation platform with FastAPI, NLP, Transformers, Hugging Face, and Streamlit—achieving 80–90% accuracy and cutting course search time by 50%.