CREDIT CARDS:1. 𝐔𝐧𝐜𝐨𝐯𝐞𝐫𝐞𝐝 𝐊𝐞𝐲 𝐃𝐫𝐢𝐯𝐞𝐫𝐬 𝐨𝐟 𝐄𝐚𝐫𝐥𝐲 𝐀𝐝𝐨𝐩𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐉𝐮𝐩𝐢𝐭𝐞𝐫 𝐑𝐮𝐩𝐚𝐲 𝐂𝐚𝐫𝐝:Led to enforcement of a new acquisition strategy which has 2x higher take rate.2. 𝐒𝐞𝐭𝐭𝐢𝐧𝐠 𝐔𝐩 𝐄𝐧𝐝-𝐭𝐨-𝐄𝐧𝐝 𝐃𝐚𝐭𝐚 𝐏𝐢𝐩𝐞𝐥𝐢𝐧𝐞𝐬 𝐚𝐧𝐝 𝐛𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐄𝐱𝐭𝐞𝐧𝐬𝐢𝐯𝐞 𝐓𝐫𝐚𝐧𝐬𝐚𝐜𝐭𝐢𝐨𝐧𝐬 𝐚𝐧𝐝 𝐁𝐢𝐥𝐥𝐢𝐧𝐠 𝐝𝐚𝐬𝐡𝐛𝐨𝐚𝐫𝐝𝐬𝐮𝐬𝐢𝐧𝐠 𝐓𝐚𝐛𝐥𝐞𝐚𝐮 𝐚𝐧𝐝 𝐃𝐁𝐓 influencing data-led decision making with:a) Transaction Success Rate (Breakdown by Transaction Life-Cycle Stages) - Improved Card Transaction Success Rate by 10ppb) Credit Limit Utilisation - Enhanced User Experience for Transaction Failures due to High Limit Utilisation.c) Transaction User Life Cycle (New, Repeat, Resurrected, Dormant)d) Credit Consumer Life Cycle (Transactor, Revolver, Delinquency)GROWTH:1. 𝐑𝐞𝐜𝐨𝐦𝐦𝐞𝐧𝐝𝐢𝐧𝐠 𝐍𝐞𝐱𝐭 𝐁𝐞𝐬𝐭 𝐂𝐫𝐨𝐬𝐬-𝐒𝐞𝐥𝐥 𝐟𝐨𝐫 𝐔𝐬𝐞𝐫 𝐋𝐢𝐟𝐞 𝐂𝐲𝐜𝐥𝐞 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭 𝐏𝐢𝐥𝐨𝐭 𝐮𝐬𝐢𝐧𝐠 𝐊-𝐌𝐞𝐚𝐧𝐬 𝐂𝐥𝐮𝐬𝐭𝐞𝐫𝐢𝐧𝐠 & 𝐏𝐫𝐨𝐩𝐞𝐧𝐬𝐢𝐭𝐲 𝐁𝐚𝐬𝐞𝐝 𝐌𝐨𝐝𝐞𝐥:25% uplift in Product Activation2. 𝐄𝐬𝐭𝐚𝐛𝐥𝐢𝐬𝐡𝐢𝐧𝐠 𝐔𝐬𝐞𝐫 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐮𝐫𝐚𝐥 𝐏𝐞𝐫𝐬𝐨𝐧𝐚𝐬 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐏𝐥𝐚𝐭𝐟𝐨𝐫𝐦3. 𝐌𝐚𝐩𝐩𝐞𝐝 𝐔𝐬𝐞𝐫 𝐚𝐜𝐭𝐢𝐯𝐢𝐭𝐲 𝐭𝐨 𝐌𝐨𝐧𝐭𝐡-1 𝐑𝐞𝐭𝐞𝐧𝐭𝐢𝐨𝐧 𝐮𝐬𝐢𝐧𝐠 𝐃𝐞𝐜𝐢𝐬𝐢𝐨𝐧 𝐓𝐫𝐞𝐞 𝐂𝐥𝐚𝐬𝐬𝐢𝐟𝐢𝐞𝐫 𝐚𝐧𝐝 𝐢𝐝𝐞𝐧𝐭𝐢𝐟𝐢𝐞𝐝 𝐊𝐞𝐲 𝐀𝐜𝐭𝐢𝐨𝐧𝐬 𝐚𝐫𝐨𝐮𝐧𝐝 𝐒𝐞𝐭𝐮𝐩, 𝐀𝐇𝐀 & 𝐇𝐚𝐛𝐢𝐭 𝐌𝐨𝐦𝐞𝐧𝐭 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤:Devised a gamification strategy with incentives that increased Month-1 Retention from 38% to45% over 3 months.4. 𝐔𝐬𝐞𝐫 𝐒𝐞𝐠𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 𝐛𝐚𝐬𝐞𝐝 𝐨𝐧 𝐑𝐅𝐌 𝐌𝐨𝐝𝐞𝐥𝐥𝐢𝐧𝐠 & 𝐀𝐩𝐩 𝐀𝐜𝐭𝐢𝐯𝐢𝐭𝐲, 𝐥𝐞𝐚𝐝𝐢𝐧𝐠 𝐭𝐨:- 60k uplift in transactions through targeted campaigns. - Increased Average Transaction per User by 4x and Spends per User by 2x. - Resurrected 10k dormant users with an average transaction frequency of ~3.5.

𝟭. 𝗘𝘅𝗽𝗹𝗼𝗿𝗮𝘁𝗼𝗿𝘆 𝗧𝗶𝗺𝗲 𝗦𝗲𝗿𝗶𝗲𝘀 𝗔𝗻𝗮𝗹𝘆𝘀𝗶𝘀 𝗳𝗼𝗿 𝗗𝗲𝗺𝗮𝗻𝗱 𝗙𝗼𝗿𝗲𝗰𝗮𝘀𝘁𝗶𝗻𝗴: Using Traditional Time Series models like SARIMA and Fb-prophet in Python for univariate time series to forecast product demand. Visualizing the series with Plotly.𝟮. 𝗜𝗻𝘃𝗼𝗶𝗰𝗲 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴: Developing bounding box algorithm with Google Cloud Vision's OCR output and built structured tabular extraction with OpenCV, which is integrated with KPMG AI Labs platform.𝟯. 𝗣𝘆𝘁𝗵𝗼𝗻 𝗦𝘂𝗽𝗽𝗼𝗿𝘁: Offering consultancy to KPMG's client for image processing and output processing tasks of invoices , in order to improve the OCR output and writing regex- based scripts to make the invoice info structured.

Worked on:1. 𝗣𝗮𝗻 𝗮𝗻𝗱 𝗔𝗮𝗱𝗵𝗮𝗮𝗿 𝗖𝗮𝗿𝗱 𝗗𝗲𝘁𝗮𝗶𝗹𝘀 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻:Automating the process of extracting information of the clients and storing it in the company’s databases using Python.2. 𝗕𝗮𝗻𝗸𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁 𝗔𝗻𝗮𝗹𝘆𝘀𝗲𝗿: Automating BankStatement document processing (converting pdfs to excel sheets) and getting valuable data-frames used for analysing the client behaviour.They could be converted to charts for more valuable insights.