Ravindra Jangir Email and Phone Number

Quantitative Model developer in Scenario Analytics

- Developing Risk models for Two wheelers, commercial vehicles and used car lending segmentBehavioral Score Model for Used Car Loan Pre-Approval• Developed a behavioral score model using PySpark, Python incorporating 12,000 model variables derived from on-us, bureau, and enquiry data. • Built Random Forest models for refinance and repurchase with Gini coefficients of 51% and 57%, ensuring stability with PSI values of 4.07 and 4.2 • Impact : reduced 23% risk with the same qualifying population.Propensity Model for New/Used Car Loan Takeup• Developed an XGBoost model with 43% Gini coefficient to predict the propensity of a 50 Lac customer base for taking up new/used car loans (on-us or off-us) using sampling techniques. Product Analytics & Other• Developed model monitoring process using cloud-era for all risk analytics portfolio.• Analyzed FastTag Data for cross-selling of vehicle loans to affluent customers.

Embarked on an unforgettable journey with 500 dynamic change makers and entrepreneurs on the renowned Jagriti Yatra. Traversed diverse landscapes, engaged with local communities, and sought the diamond of India. Leveraged this immersive experience to hone leadership skills, foster innovation, and drive positive change. Grateful for the opportunity to contribute towards a brighter future for our nation.

– Led a three-member team to manage entire Retail and Corporate Trade (international & domestic) analytics portfolio– Wallet share calculation: Used string matching technique (SQL, Python) with external shipment data to calculate wallet share and its movement for existing clients to identify cross-sell and new to bank opportunities.Impact: 44 customers with 0% wallet share routed 12 billion worth of bills and 50 new customers acquired– Defined project roadmap, managed data downloading and injection, coordinated with stakeholders, and created a value-bucketed wallet share report. Mapped opportunities in relationship manager dashboards for enhanced visibility.– Developed Prediction model (python, SQL) using Naive Bayes classifier for identifying transactions as Bank or Customer induced. Impact: funds of ∼ 400 accounts amounting 4lac were saved every month from transferring in RBI DEAF accounts– Developed SMS Campaign for customers whose accounts are likely to be turned DEAF in next 6,3,1 months (SQL)– Developed and launched the first four ”Direct to Customer” digital campaigns on CIB(Corporate Internet Banking) using clickable banners to drive engagement. Devised an efficient SQL script featuring prioritization and deletion logic to enhance data management.– Generated revenue of ∼ 10 Mn by using text mining of SFMS message, string matching and Round Robin Methodto develop strategies for 360 coverage of LC-BG payments (SQL, Python, SAS)– Generated revenue of ∼ 40 Mn, Developed 10 campaigns by identifying top profit/loss-making profiles and recommend product for profit maximization & developed algorithm to assign leads to branch officers (SQL)– Developed Zone-wise Campaign-Relationship manager effectiveness report/MIS for all live campaigns in trade(SQL,SAS)

- To strategize the content curation for social media to enhance user engagement of Quriverse, a startup project- Tracked daily active users in Google Analytics and identified KPIs to separate out different tranches of moderators- Used SQL, Wilson score and Bayesian statistics to Built a model to measure social score of user on the platform

- Analysed dataset of FCCU and developed material balance and heat Balance models to calculate coke yield- Got a recommendation letter for higher studies for excellent work done during the internship