k**************e

Working with LLMs ● Worked team collaboration with some vibrant colleagues on the Data Science and Artificial Intelligence(AI) team ● Worked with LLMs for dataset generation for training and fine-tuning LLMs like the mistral, Llama2, and other LLMs ● Fine-tuned LLMs using the PEFT technique for blog content generation, and text summarization ● Made use of google-collab, Paperspace, and the AWS-Sagemaker (Cloud Platform) to fine-tune the LLMS ● Adopted the huggingface platform for the collaboration, model deployment to inference endpoint, and communicated the endpoint to the next user Worked on LLM applications ● Developed a proactive Cross-Functional document reader product using Langchain frameworks and the OpenAI GPT using the Retrieval Augmented Generation which enhances Artificial Intelligence usage ● The system can connect to different sources like Excel, PDFs, CSVs, word documents, and other sources to retrieve information for non-technical users ● Made use of MongoDB database for data storage ● Deployed the solution as product into a chat interface system to query the document

Collaboratively gathered the dataset, conducted exploratory and trend analysis, cleaned the data, and applied data mining and data analysis techniques to extract key insights ● Developed 16 machine learning models and Statistical Modeling for real estate price valuation with a mean absolute error of 012 and an R² score of 0.89 (87%) in performance ● Conducted comparative analysis of models to select the best-performing one based on key metrics, and created visualization plots to present the results for clear, visual analysis. Conducted hyper-parameter tuning for model optimization, achieving a minimum accuracy of 89% in evaluation ● Created a dashboard to communicate effectively the result to stakeholders ● Deployed the model on render for testing and usage

Worked with and supervised a vibrant team of 12 members ● Research, extract and compile data from various organizations, including universities, secondary schools, and entrepreneurs, spanning about 15 different domain ● Gather vital information such as email addresses, phone numbers contacts, company names, and ratings from approximately 500 organizations in each task using web scraping techniques ● Maintain data integrity and security measures to protect sensitive information ● Utilize data cleaning techniques to ensure accuracy, completeness, and suitability for use ● Collaborate with team members to streamline data collection processes and improve data quality ● Communicated the result of to the stakeholders and management in a non-technical term