Rishikesh Sahoo

• Built machine learning models to segment customers by behavior and upgrade likelihood, improving targeting precision and boosting engagement by 20%.
• Applied predictive analytics to optimize campaign timing and messaging, enhancing performance and conversion rates.
• Developed LLM-based Campaign File Audit Validator using Claude Sonnet 3.7 on DataBricks, ensuring technical specifications from business requirements.
• Created AI Chatbot with Co-pilot Studio for team-wide knowledge base, supporting process queries and multiple functionalities.
• Utilized SAS, SQL, and Python to analyze credit card activity, translating insights into data models and DAX measures for key risk indicators.
• Participated in requirement meetings with project owners and US stakeholders, driving data-driven campaign strategies.

• Collected and transformed loan applications and customer data using SAS, SQL, and Python, automating daily/weekly/monthly BAU and MIS reports.

• Extracted and analyzed credit/debit card usage from merchant transactions, creating summary datasets to monitor behaviors efficiently.

• Designed end-to-end process automation integrating Python outputs with Excel and Power BI dashboards for advanced visualization and KPIs.

• Executed logistic regression projects in Python for credit risk modeling, focusing on key factors with SQL for data cleaning and preprocessing.

• Transitioned reporting from Excel/PowerPoint to dynamic Power BI storytelling dashboards, improving quick summary analysis.

• Developed credit risk assessment models, reducing loan defaults by 3-4% through analysis of 300,000+ customer datasets.

• Identified and reevaluated loan distribution patterns, reducing repayment risk by 14% via data-driven strategies.

• Collaborated with cross-functional teams to improve credit scoring models, increasing accuracy by 10%.

• Promoted data-driven decision-making culture, boosting data-based decisions by 20% across the organization.