GenAI-Powered Assistant (RAG & Fine-Tuning)Developed an AI assistant leveraging Retrieval-Augmented Generation (RAG) and fine-tuned LLMs to automate query resolution for email and chat based user queries.Designed a pipeline integrating a vector database (Pinecone) for semantic search and retrieval of relevant email and chat data.Fine-tuned GPT-4 using domain-specific datasets to enhance accuracy in summarization, prioritization, and professional tone adaptation.Built a scalable backend using FastAPI, integrated with Microsoft Graph API for email and calendar access.Deployed a robust solution with Docker and Kubernetes on Azure, ensuring low latency and high scalability.Achieved 95% accuracy in context-grounded responses, reducing workload by 50% and improving task prioritization efficiency.Technologies: Python, GenAI, LangChain, OpenAI GPT-4, Microsoft Graph API, Pinecone, Docker, Kubernetes, Azure

Assisting Prof. Yiwen Gao with research in Cybersecurity.Utilised advanced web scraping techniques to gather data from diverse sources and used the fuzzywuzzy library to identify and analyse similarities within datasets.Automated the data scraping process to enhance efficiency and accuracy, demonstrating strong problem-solving and technical skills in support of research objectives.Scraping the required companies data from various sources and finding similarities using fuzzywuzzy library.Evaluated and assessed student performance in CIS-370 (Business Analytics) course, employing robust grading methodologies to measure comprehension and application of course materials.Tutoring the students in CIS department and clarifying the concepts that were taught in various courses in College of Business department.

Project: Early Cancer Detection in Adults (Medical Science)Responsibilities:Implemented blockchain and artificial intelligence technologies to develop a platform facilitating early cancer detection in adults, with a specific focus on the youth age group.Collaborated with a multidisciplinary team to create a system connecting patients and healthcare providers with on-demand fellowship-trained expert radiologists globally.Utilised AI algorithms to assess the likelihood and risk of developing cancer based on patient symptoms, leveraging a comprehensive database of diseases and past symptoms.Employed feature engineering that led to the creation of a robust dataset comprising 1,000 features, ensuring data integrity and relevance.Developed and implemented an initial-level model based on patient symptoms, contributing to early cancer detection efforts.Collaborated with stakeholders to continuously refine and improve the platform's functionality and performance.Project : Natural Language Processing (NLP) and Sentiment Analysis ProjectResponsibilities :Conducted sentiment analysis on youth sentiments by web scraping data from Twitter, Instagram, and UNICEF, utilising Python.Performed Exploratory Data Analysis (EDA) to uncover patterns and distributions in the collected data.Employed advanced visualisation techniques using Plotly to create interactive basemaps, word clouds, violin plots, scatter plots, and dynamic heatmaps, facilitating deeper understanding of youth thoughts and sentiments.Visualised sentiment data over geographic locations using Tableau to provide insights into regional sentiment variations.Utilised Latent Dirichlet Allocation (LDA) to identify key topics in the data, revealing prevalent themes in youth sentiments.Technologies used : Python (NumPy, pandas, scikit-learn, TensorFlow, PyTorch), Plotly, Tableau, Latent Dirichlet Allocation (LDA), AWS Sagemaker

Developed an end-to-end system for vendors using REST APIs so as to enable online casino facilities.Successfully integrated new vendors and games for njord ventures’ all brands so as to facilitate gaming operations for its customers across various countries in Europe.Enhanced the user registration and regulatory services to enable regulatory compliance specific to countries so as to launch the operations of the client in those countries.Improved the cashier service so as to enable cryptocurrency transactions for the njord customers.Designed and implemented a frontend website for vendors to conduct online casino operations to deliver an intuitive user experience.Diligently adhered to Agile methodologies and utilised the Scrum framework on a daily basis to effectively manage and deliver projects within designated timelines.Followed CI/CD pipeline techniques using git and jira to enhance project execution and team collaboration.

Led the development of a Natural Language Processing (NLP) model aimed at extracting key skills from diverse job postings.This model laid the foundation to the creation of an innovative application, empowering students and government entities to pinpoint crucial industry skills for tailored training strategies.Significantly boosted accuracy to an impressive 90% by integrating a Named Entity Recognition (NER) model, demonstrating a commitment to precision and excellence in skill identification.

Initiated and executed a project focused on data acquisition from diverse local businesses across tier-2 and tier-3 cities as well as small towns, utilising multiple websites as data sources.Spearheaded the development of a comprehensive application offering centralised access to business information across various locales.Implemented a Python-based web crawler tool to automate data retrieval processes, significantly enhancing efficiency and reducing scraping time for diverse website formats.