Mallela Vema

I've cleaned and analyzed 50,000+ records with Python and SQL, improving data quality by about 30% and reducing manual reporting time by 40%.

During my Data Science & AI internship at Besant Technologies, I built interactive Power BI and Tableau dashboards for sales, profit, and customer behavior, helping stakeholders act on insights twice as fast. I also applied regression and classification models with scikit-learn, reaching approximately 85% prediction accuracy on historical data.

I've extended my analytics work into GenAI by designing a FAISS-based RAG pipeline, a CrewAI multi-agent content pipeline, and a FastAPI document-generation service. My Voice AI Customer Support Agent combines LiveKit, LangGraph, RAG, ChromaDB, Deepgram, Groq, and Cartesia for real-time customer support.

I also build practical data products, including healthcare disease prediction analysis, an automated Amazon sales dashboard, and a Netflix content intelligence dashboard. I enjoy translating business questions into clean data, useful KPIs, and actionable visual insights.