Juhi Roshan

I am an analytical, detail-oriented data science professional with hands-on experience building and deploying machine learning models and developing automated evaluation workflows.

During a Data Science internship at Larsen & Toubro, I deployed models with Docker and Kubernetes, integrated Prometheus and Grafana for monitoring, and created Power BI dashboards to visualize model performance for stakeholders.

I have implemented generative AI prototypes including LLM fine-tuning, chatbot development, RAG pipelines, and BERT-based summarization, and I built a real-time support assistant using LangChain, OpenAI API, FAISS, and Streamlit.

With an M.Tech in Computer Science and prior experience in data validation and quality engineering at Accenture, I combine strong programming (Python, SQL, Java), deep learning, and cloud-native tooling to deliver reliable, scalable AI solutions.