LAKESH KUMAR SURYADEVARA

Built a dynamic Power BI dashboard to monitor sales and profit across 230+ cities. Collaborated with users to define metrics and develop executive-level visualizations. Integrated data from SQL and flat files into a centralized model for analysis.

Developed a Power BI dashboard for key financial KPIs, cutting manual reporting by 40%. Worked with stakeholders to meet strategic needs and analyzed trends to provide actionable recommendations that influenced planning discussions.

Created a Power BI dashboard for live crypto market data. Automated data ingestion and cleaning with Python and SQL to ensure accuracy. The dashboard enabled data-driven investment discussions by simplifying complex market trends for stakeholders.

Developed a complete system for live monitoring of data center health and energy use. This includes gathering sensor data, processing it on the fly for metrics like PUE, finding anomalies, and showing everything on a web dashboard.

Developed a robust time-series forecasting model to predict crop yields under varying climate conditions. Utilized historical weather and agricultural data to provide actionable insights for farmers and policymakers, enhancing food security strategies.

Implemented a deep learning solution for classifying brain tumors from MRI images. Focused on model accuracy, interpretability (using techniques like Grad-CAM), and efficient deployment for potential clinical application.

Developed a forecasting system for urban energy consumption, incorporating exogenous variables and handling data anomalies to provide more robust and reliable predictions for grid optimization and smart city planning. Included a Streamlit dashboard for interactive visualization.

Created a web dashboard that performs real-time sentiment analysis on financial news articles using a fine-tuned BERT model (FinBERT). Provides instant insights into market sentiment, aiding investment decisions. Integrated with NewsAPI for data retrieval and Plotly for interactive visualizations.

Developed an automated deals agent to track product prices across e-commerce sites and predict optimal buying times, especially for seasonal sales like Black Friday. Users receive SMS notifications via Twilio when prices drop significantly, based on historical data analysis.

Designed a comprehensive, robust, and scalable database architecture for Vendohub, an MLM e-commerce platform. The project emphasizes data integrity, security (including encrypted sensitive data and unique constraints), and comprehensive functionality, highlighted through a detailed ER Diagram.