• 🌍 Based in Hyderabad, India — open to US Remote roles
• 📊 I turn raw data into clear business insights through dashboards, reports, and predictive models
• 🛠️ Experienced in Python, SQL, Power BI, and Excel across finance, operations, and ML domains
• 🎓 Master's in Computer Science — Adikavi Nannaya University (2025)
• 🎯 Seeking entry-level Data Analyst / Business Analyst positions

Tools: Python, scikit-learn, Pandas, Matplotlib

• Built a logistic regression classifier on 284,000+ transactions with SMOTE to handle class imbalance
• Achieved 87% accuracy & 94% AUC-ROC score, reducing false negatives vs baseline
• Visualized confusion matrix and ROC curves for non-technical stakeholders

Tools: SQL, MySQL, Power BI

• Queried 10,000+ order records using CTEs, window functions, and JOINs to find top SKUs and peak hours
• Found top 3 categories drove 65%+ of revenue; Friday evenings = 22% of weekly orders
• Delivered structured SQL scripts and PDF report covering revenue trends and category breakdown

Tools: Power BI, DAX, Data Modeling

• Built end-to-end interactive Power BI dashboard analyzing credit card spending patterns across 4,000 customers in 5 cities
• Identified high-spending customer segments and peak payment periods using DAX measures and dynamic slicers
• Delivered executive-ready visuals supporting data-driven credit card product strategy

Tools: Python, Random Forest, scikit-learn, Pandas

• Built Random Forest regression model on EPA data predicting AQI from 5 pollutants — achieved R² = 0.95 & MSE = 0.058
• Performed end-to-end EDA, feature engineering, and StandardScaler normalization
• Feature importance analysis revealed pollutant mean concentration as strongest AQI predictor

• 📧 Email: devikalaharibandi@gmail.com
• 💼 LinkedIn: linkedin.com/in/devika-lahari
• 🐙 GitHub: github.com/Devikalahari03