Developed and implemented a high-performing multi-class text classification model using advanced NLP techniques,
including BERT, Sentence Transformers, and SetFit Models, resulting in 85% overall accuracy in accurately predicting
and classifying various industries within the banking sector.
Successfully delivered a business intelligence solution utilizing SAS VA for report development and SAS Data Integration
Studio for efficient data flow processes to enhance reporting capabilities for the network of authorized repairers
(partners). Key achievements include streamlining data management, processing, and generating comprehensive
reports using SAS VA.
Developed a highly efficient Chatbot using the Rasa framework and Python, which significantly reduced search time for
information retrieval. The Chatbot utilizes natural language processing to accurately detect intent, automate SQL
queries on a database, and extract relevant information from PDF documents. The Chatbot's capabilities improved
processes and streamlined workflow, exemplifying expertise in leveraging advanced technologies to drive efficiency and
productivity.
Developed an Intelligent Solution for Smart Banking leveraging machine learning and artificial intelligence capabilities.
Key highlights include customer segmentation through clustering, integration of a Recommendation System for
targeted product assignment, and real-time dashboard for data-driven decision making by top management.
Successfully enhanced the overall customer experience and achieved significant results, with a remarkable
improvement of nearly 20% in the bank's product adoption rates within the first three months of deployment. This
project showcased expertise in Python, Django, Flask, Power BI, HTML, CSS, and SQL Server Database.
Implemented a machine learning-based Credit Risk Model for predicting probability of default (PD) and lost given
default (LGD) using logistic regression. Led the data collection, cleaning, and processing efforts, including feature
engineering, to improve model performance. Utilized these predictions to calculate expected credit loss (ECL) for the
organization. Improved overall credit risk analysis and decision-making, resulting in successful outcomes.
Successfully executed web scraping from different sources and data extraction from PDF using OCR technics and
Implemented batch jobs to automate data collection and storage process.
Developed and trained machine/deep learning models for multi-class text classification with an accuracy of 81 %.
Improved the classification performance by 5 % through feature engineering and model tuning.
Implemented recommendation systems using different approaches such as content-based, collaborative filtering, and
hybrid methods.
Designed and implemented a Chatbot/VoiceBot using Rasa, NLP, NLU, text-to-speech, and speech-to-text techniques.
Created APIs using FastAPI to enable seamless interaction between users and the platform.