Irshad Mujawar

Anomaly Detection Models: Trained and deployed a number of
anomaly detection models; stand-alone and ensembles.
MongoDB Pipelines: Built the data pipelines which consume data
from HIVE and pushes on MongoDB Servers.
Reporting: For various processes within the organization, created automated reports for getting alerts for anomalous data.

Senior Data Scientist
Pelatro Solutions
04/2019 - 07/2022, Bengaluru, India
Propensity Models: Churn, Change in Usage/Spend Models building and
deployment.
Recommendation Engine for Campaigns: Given a subscriber
segment, predict the best possible offer/s.
Reinforcement Learning (RL): Used RL Tabular Methods for reducing
the inactivity in the customer base.
Bayesian Networks: To determine the cause of the churn.
Ideation and Collaboration with Business Team: Helped to
determine the roadmap for analytics in the product.
CLV: Using Markov Models built Customer lifetime value models.

Data Scientist
Pelatro Solutions
01/2017 - 03/2019,
DT Models and Insights: Built the tree-based models to predict the
fraudulent MSISDN purchases with high accuracy.
SOM Models (Customer Segmentation): Built a SOM model to break
the subscriber base into smaller segments.
Report Automation: For Large files (> 10GBs compressed).