Churn uplift model and ROI simulator
Targets retention offers at customers the offer actually changes, instead of those most likely to leave.
Open to graduate roles, internships and freelance work · UK and EuropeStudent visa · sponsorship required
MSc Data science student working on uplift modelling, forecasting and computer vision in Python and SQL. I build models that end in a decision and measure the result in pounds.
Targets retention offers at customers the offer actually changes, instead of those most likely to leave.
Forecasts TfL rail electricity two years ahead from TfL, GLA and Met Office data, with cost and CO₂ in an interactive app.
Classifies congestion from camera imagery using a CNN with YOLOv8 vehicle detection. Built at CSRTC.
Sales forecasting for inventory decisions, and an ANPR system that detects and reads plates.
Machine Learning, Big Data, Data Visualisation. Researching LLMs for scoring and generating survey responses.
Built the CNN and YOLOv8 congestion pipeline and lifted accuracy by 14% through augmentation and preprocessing.
Built a stock management app on MySQL with SQL reporting and validation that cut manual reconciliation.
GPA 3.6 / 4.0
Graduate roles from Feb 2027 (sponsorship required) · Internships now · Freelance projects
krishitajbani@gmail.com