Muhammad Mudasser Afzal

- Improved the automated resume parsing process to extract relevant developer details and generate better embeddings. - Feature engineering on developers' data includes; generating new features using the baselines. This process helps us to improve the pairwise ranking score by 2%. - Implemented Automated Pipelines for data fetching and modeling on vertexAI using BigQuery & Python.

 Created Machine Learning modules i.e., forecasting, anomaly detection, etc. Built & maintained the backend of clustering and Data Engineering modules i.e Algo plus.  Worked on the lost sale algorithm for Vodafone to monitor and optimize their inventory and in-stock items across 500+ stores.  Analyzed data with more than 20 million records to drive optimization and improvement of product development, marketing techniques, and business strategies

 Developed solutions for a call center working on sentiment analysis, NER, intent classification, and multi-label classification getting more than 80% of accuracy on each task.  Implemented a Recommendation System for recommendations of brands using GNN & Hybrid Recommendation techniques which retrieves 80% of top relevant brands.  Led a team of 3 ML Engineers for the development of ML projects.  Added value to the organization by retaining long-term clients and increasing the company’s revenue by 45%.