India

Data Scientist

AI/ML Engineering

Data Analysis

Bachelor of Technology (B.Tech.)

Gaurab Das

Lead Data Scientist

Communication

Collaboration

Problem-solving

Time management

Critical-thinking

Stakeholder management

AI/ML

Machine learning

Mathematics

Python

Introduce yourself, define mentoring expectations and goals. Learn and grow together.

General mentorship

Building analytical models and solutions for customer lifecycle management.

Corridor Platforms - An AI driven decision workflow governance and automation capability to rapidly transform risk and marketing decisioning at lending organizations for effective revenue growth, risk control and operations.

• Developing capabilities for decision making & analytical framework.

• Designing use-cases like Credit Line Increase (Probability of Default, Balance Prediction Models) & Text analysis (NLP Topic Models) for the Platform.

• Coordinating with clients to fast track porting of risk models on the Platform for real time and near real time decision making.

Key Learnings : Automation | Structured thinking | Articulation

AvailFinance - A Neo Bank for the Indian Blue Collared workforce!

• Risk Scorecards: Identified high risk profiles by leveraging Traditional & Alternative data sources & by using ML based algorithms such as XGBoost, GBM, LGBM, Random Forest & Logistic Regression. 
        1. Application  Scorecard: Increased approval rates by 40% as well as decreased default rates by 13% for a major customer segment.
        2. Behavioral Scorecard: Used the information of accounts on books  & modelled the different behavioral features to measure the risk of a customer on the upcoming EMIs.

• Marketing Problems: Identified  segments for Targeted marketing using ML algorithms. 
        1. Referral Segmentation: Identified a segment using Decision Tree which   will bring in good quality referrals to the business.
        2. Customer Propensity: Classification  model to identify accounts at the top of the funnel, that would complete the documentation process in the loan application to increase conversions.

• Planned & developed an Alternative Data Solution System that serves as an essential counterpart to Traditional Data for building a Social & Financial profile. Extracted features from various data sources like sms, geolocations, web scraped data, installed apps etc., to make credit assessment simple. 

• Credit Bureau Switch: Executed end to end analysis, comparison & development of variables based on two different bureaus being used in strategies, scorecards & policies resulted in saving 75% cost per loan disbursal.

Key learnings : Lending | Collaboration | Leadership essentials

To fit an appropriate time series model (Model Specification and Diagnostics) and forecast Consumer Price Index.
Software Used: R

Ugam - a leading analytics company catering across industries like Retail & Consumer Brands.

Brand awareness for a leading US retailer using Path Analysis (SEM)
•   Identified important KPI’s for the director of the company using Path Analysis models which helped understand the hypothesized causal connections of the KPI’s on brand visibility and, proposed 20% scope of improvement across various geographies.
•   Built simulators with the identified KPI’s using Excel-VBA

Awards:
Certificate of Recognition - June 2018

Product Matching Engine for a leading US retailer:
•  Developed strategies and automated a manual quality control process for an e-commerce giant using Python resulting in reduction of 100 hours per week of manual reporting time
•  Engaged in text & image classification using various machine learning techniques like Word2vec, Opencv, Inception V3, ResNet etc to build comparison models
• Gap analysis for the manual quality control process


Awards:
Certificate of Recognition - March 2018

SAP-ABAP developer. Worked on Analysis, Evaluation and Implementation of defects raised at the user level for a German Automotive Giant. The business focused on resolving issues for central financial template.

Pitchtion - a bootstrapped market place for simplified money lending and borrowing system by defining a credit worthiness engine for the Micro, Small and Medium Enterprises. The engine would be an effective credit scoring system which will be used by NBFCs to determine the worthiness of a loan application .

Gaurab Das🇮🇳

About Gaurab Das

Expertise

Sessions with Gaurab Das

What mentees say