Sanchit Mittal

Merchant Targeting Optimization:

• Analyzed transaction, demographic, and service usage data to provide actionable insights for improving task efficiency for on-ground teams
• Automated the whole method to provide real-time task generation based on various business variables which helps in direct targeting of high-potential merchants, significantly improving merchant acquisition rates

Fraud Detection and Prevention:

• Identified unethical activities by agents that led to mis-selling and resource wastage, provided with insights and methods preventing further losses in assets and payments
• Developed data-driven strategies to optimize agent management and minimize fraud-related inefficiencies

Agent Performance Tracking:

• Analyzed agent activity data to identify high-performing and low-performing agents, enhancing tracking and operational efficiency
• Delivered insights that improved resource allocation and boosted the overall productivity of agent teams.

Search Product Analyst:

• Analyzed search and install patterns, providing insights that improved search functionality and increased app installations
• Collaborated with product teams to enhance the user experience, leading to better search design and user satisfaction increasing installs by 38%

Data Model Design:

• Spearheaded the design and implementation of a new data capturing model for the App Bazaar revamp project, improving data flow and collection.

Lead to Order:

• Built a propensity model for the sales team of a fortune 10 technology giant to predict the likelihood of an opportunity to convert to an early sales stage
• Leveraged a hypothesis driven approach to identify the statistically significant features affecting the likelihood of conversion
• Enabled the team to prioritize leads based on multiple factors which were fed into various classification algorithms attaining model accuracy of 78% with potential cost savings of 80M

Days to Close:

• Built a classification model for the sales team of a fortune 10 technology giant to predict the period of days required to convert a lead to an order
• Built upon the previously built L20 project using the classification method to predict a range of period of days required to close the deal
• Led to realization about the data manipulation performed by the sales team to reduce days to close.