Shivam Aggarwal

About Shivam Aggarwal

Data engineering and analytics professional with a focus on developing data pipelines, backend servers, and risk detection frameworks. Experienced in enhancing daily activity tracking, streamlining machine learning model deployment, and designing fraud detection triggers. Proficient in Python, Spark, AWS, Flask, and SQL. Passionate about playing basketball and video games in my free time.

Experience

1. Risk Analyst

   Finnable

   Oct 2023 — Present · Bengaluru, IN

   Developed Pipeline for Daily MIS:Objective: Enhanced daily activity tracking across multiple metrics including Collection/Underwriter Efficiency, Sales, and Portfolio Tracking.Outcome: Facilitated easy tracking of team performance for managers.Tech: Pulled data from AWS Athena, performed calculations and transformations in Python (Pandas/Spark), sent daily emails, and scheduled pipelines on a daily/weekly/monthly basis.2. Created Backend Server Using Flask:Objective: Streamlined the deployment of diverse machine learning models and optimised the customer journey for the lending process.Outcome: Simplified model deployment and expedited customer journey processes.Tech: Created backend server through Flask API, used AWS EC2 (AMI) for server deployment.3. Risk Triggers Across DSA\'s, RM\'s, and Verification Agencies to detect Potential Fraud:Objective: Identify potential fraud across different metrics of the DSA portfolio.Key Contributions: Designed a framework for an automated monthly trigger report to identify risky RMs, DSAs, or verification agencies before they exceed portfolio benchmarks.Metrics Monitored: PF Yes approval rate, RV/OV approval rate, bank/company penetration, and business volume increase.4. Address and Company Match Logic Development for Data Verification:Objective: Provide RV waiver to customers based on address verification.Key Contributions: Developed Python-based logic using Levenshtein ratio technique for address and company name matching, resulting in a 35%-40% increase in cases processed by Speedy.5. Development and Management of New to Credit (NTC) Policy:Objective: Increase approval rate for DSA customers by ~6%.Key Contributions: Coded and implemented NTC policy in Python, conducted comprehensive testing, and managed backend operations ensuring prompt customer approvals.

Education

• Panjab University

  Bachelor of Engineering - BE

  2018 — 2022