Selected Projects

Next Equipment repair Prediction

Objective: ML based solution that predicts the next equipment repair based on the past repairs.
Results:
• Developed a tree-based regression model with confidence intervals (upper and lower bounds) for the prediction.
• The model predicts the next equipment repair (+/- 30 days) with 75% accuracy.

Ticket Automator Solution (TAS)

Objective: ML based solution to bring meaningful insights on the CRM tickets thereby reducing the time and effort spent by the customer service agents, increasing the customer digital experience.
Awards:
• Won 3rd prize for the TAS solution in the YesBank global Datathon Sept-Dec 2018.
• Received citations of the POC in a few technical blogs.

Which celebrity do you resemble ?

What it does: Given an input image, the Deep Learning model compares among 33 different celebrities and displays the top 3 celebrity resemblance.
Results:
• Predits with 86% accuracy the exact celebrity.
• Accuracy compared to few implementations posted online.

Intelligent Data Extractor and Analyzer (IDEA)

Objective: ML and NLP based solution designed to reduce manual effors by extracting useful infomation from the PDF.

IDEA has modules for classifying blocks into different domains, extracting tables, summarizing text and parsing dependencies on PDF files.