About Pawan Lamba

A Big-Data Machine learning Data Scientist experienced in domains like NLP, Banking, Marketing, Comparison Shopping, SEM, Channel and Pricing Optimization. Competitive in various tools and Skills including

- Deep Learning - Machine Learning - Revenue Optimization
- Distributed Deep Learning - Adaptive Pricing - Statistical and Quantitative Modeling
- Scala & Python, TensorFlow - Classification and Clustering - Risk Modeling
- Spark & Big Data - Large Scale Systems - Real Time Serving

Professional Experience

Paytm - Noida India
Dy General Manager, Data Science, Collections Lending, Aug 2021 – Present
Senior Manager, Data Science, Collections Lending, Aug 2020 – July 2021

  • Built a Team of 14 comprising of Data Engineers, Analyst and Machine Learning Scientist.
  • Planned and executed, roadmap of ML and Data infrastructure for debt collections.
  • Instrumental in designing strategy with collection heads for a BNPL product.
  • Spearheaded Reinforcement Learning based solution for automated strategy driving decisions for Channels (SMS, OBD, BOT, Call, Push, Chat, Banners, Ads), Allocation , Intensity, Time of Contact, Field, Agent, Content and Language for 3rd party debt collection.
  • Built scalable ML infra for multiple lenders and loan products powering collections SAAS product.
  • Designed ML based Segmentation model pipelines for Pre-Delinquency and Roll-forward pre-GCL buckets.

Senior Manager, Data Science, CDST, July 2019 – July 2020

  • Mentor and solution a AI Chatbot project with deep learning based intent identification.
  • Entities and Intent resolution using** Deep-NLP** based models for real time customers.
  • Built Semantic Role Labelling, Language Identification and Sentiment Analysis for the Chatbot.
  • Exploring Reinforcement learning based AI dialogue manager.

Senior Manager, Data Science, Paytm Labs Team, July 2016 – July 2019

  • Built Deep Learning based search token classifier to amplify customer intent based search results.
  • Design/Model Customer Scoring Engine assessing customers on 500+ parameters. Enabling Business target/cater right set of customers.
  • Worked on the profiling of the customers across categories and interaction levels, to design an in-house Customer Lifecycle Management tool.
  • Developed LSTM based Deep Learning models to cross predict the sales of similar products.
  • Designed RNN and LSTM based translation and transliteration system for the cross language Indic text.

MakeMyTrip - Gurgaon India
Deputy Manager, Data Science , Mar 2015 – July 2016

  • Optimized the risk exposure for the hotel price cache using survival analysis , developing patented tech.
  • Modeled Booking affinity using variables based on competitive pricing, market share, intent etc.
  • Developed auto-optimizing Hadoop Machine Learning system for hotel exposure control at affiliate channel, maximizing ROI using the dynamic pricing and discrete control for 1.9+ Billion leads.
  • Detect Sale-fare using the model and density based outlier detection system.
  • Built BigData models to predict demand risk and airline price risk , and hence maximized revenue.

91mobiles - Gurgaon India
Lead Data Scientist, June 2014 – Mar 2015

  • Score for the devices using the Techno-Analytical variable set.
  • Latent Choice Modelling based on the expert choice/rankings.
  • Decompose the driver of benchmarks for the mobile/ laptop/ tablets.
  • Metric to gauge the Brand Awareness, Intent, Preference on the Comparison Shopping Space.

Shopzilla - Gurgaon India
Senior Data Scientist, June 2013 – June 2014

  • Built Revenue Maximization Prediction Models for Search Engine Bidding.
  • Adaptive Pricing Models to maximize the Net Profit.
  • Identify opportunities by predicting revenue generating keywords.
  • Building Scalable HADOOP Based Automated Big Data Modeling and Scoring Engine.
  • Social Network Analysis of Keyword Clusters to identify similarity pattern.
  • Integration of new big data technologies like PIG, HIVE HADOOP, R.

Global Analytics India Pvt. Ltd - New Delhi, India (Jan 2013 – May 2013)
Global Analytics India Pvt. Ltd - Chennai, India (Feb 2011 – Jan 2013)
Senior Analyst, Feb 2011 – May- 2013

Served as a key member of underwriting team for a loan portfolio of an UK based payday lending firm.

  • Analyze data for Underwriting the lending product in UK
  • Regression Analysis to find out predictor variables.
  • Build Statistical Credit Scoring models to predict the default rate.
  • Optimize portfolio risk with dynamic risk variables based on Application Input.
  • Code and generate variables to be used as inputs for various statistical analyses.
  • Maintain the implemented packages.

Aim Consulting - Gurgaon India
Associate Consultant, Feb 2010 – Feb 2011

Responsible for

  • Mathematical Operation Research Algorithm development and Optimization recipes for optimal usage using MATLAB, MS Excel and MS Access.
  • Design data collection modules using Excel and Access.
  • Developing and implementing Optimization Algorithms for finding the optimal production plan
  • Analytics and statistical support to the on-site project manager for the production improvements
  • Variance and Covariance analysis of the time series production data
  • To analyze the root cause of the losses in the performance.
  • Drafting Reports and presenting the key Analytical outputs to the team and the timely briefing to the end customer management on tools and usage

Kiran Udyog Pvt Ltd - Gurgaon India
Executive Operations, Nov 2009 – Feb 2010

Responsible for assisting Chief Operating Officer of the company for regular plant operations of the company. Involved in the Vendor Development process for outsourcing and client management in co-ordination with the marketing team

Education

  • M.Tech (Control & Computing-Electrical Engg), Indian Institute of Technology-Bombay(IITB), 2009
  • B.E (Electronics and Instrumentaion), ITM-Gurgaon, 2003 - 2007
  • Intermediate/+2, Raman Munjal Vidya Mandir-Gurgaon, 2001 – 2003
  • Matriculation, Raman Munjal Vidya Mandir-Gurgaon, 2001

Academic Project

  • Risk Sensitive Optimal Control –Application to Financial Optimization Surveyed and Designed algorithms for calculation of Risk measures like Mean variance, Value at Risk ( VaR ), Conditional Value at Risk ( CVaR ) and Mean-CVaR. Applied optimal control theories to the Risk Optimization in the Financial markets and came up with a dynamical model for the Market Risk.
    Finally developed a optimization algorithm using Stochastic Programming method to find optimal weights for the portfolio Asset Allocation model.
  • Speech-Music Discrimination – Built variables for classification using Audio Signal Analysis and Time series data.
  • The Linear Algebra Behind Google’s Page Rank Algorithm: Implemented and simulated the Google’s Page rank algorithm using the concepts of Linear algebra and analyzed it for various web page ranking.
  • Premature Ventricular Beat Detection using Wavelets – Modeled variables and classifier to predict probability of Arrhythmia attack.
  • Course Work : Digital Signal Processing, Wavelets, Matrix Computations, Applied Linear Algebra, Optimization, Optimal Control, Non-Linear Dynamical Systems and Multivariable Control System, Embedded systems.