01
PWC India · February 2026
GenAI-Powered Sales Analytics Chatbot
Data preprocessing, SQL, KPI tracking & automated reporting
- Designed a Streamlit analytics application in Python to track retail KPIs including multi-year revenue trends and quarterly growth.
- Used Pandas to extract, clean, and transform raw Excel data, loading results into SQLite for efficient querying.
- Built an insight engine translating SQL results into executive-friendly summaries with downloadable CSV exports.
- Integrated OpenAI API enabling non-technical stakeholders to query complex databases in plain English.
PythonPandasSQLiteStreamlitOpenAI APISQL
02
PWC India · February 2026
AI Fraud Analysis & Monitoring Tool
EDA, anomaly detection, visualization & stakeholder communication
- Analyzed daily banking transactions using Scikit-Learn (Isolation Forest) to isolate the top 3% of high-risk fraud cases.
- Designed scatter plots and charts using Seaborn and Matplotlib to highlight suspicious transaction patterns.
- Built a Streamlit dashboard with dynamic filtering to support analyst review of fraudulent records.
- Leveraged Google Gemini API to auto-generate business explanations for flagged transactions, accelerating manual review.
Scikit-LearnSeabornMatplotlibStreamlitGemini APIAnomaly Detection
03
Snapdeal · March 2026
Customer Segmentation with K-Means Clustering
Data cleaning, exploratory analysis, customer behavior & clustering
- Engineered a K-means clustering model in Python to segment customers and identify distinct purchasing behaviors for targeted marketing.
- Designed data cleaning workflows to handle missing values, outliers, and normalization for ML-ready inputs.
- Analyzed large-scale e-commerce transaction datasets to translate complex data into actionable business metrics.
PythonK-MeansPandasEDANumPyScikit-Learn
