PDF Chat Bot: Document Based Conversations
This research explores the potential of pre-trained open-source Large Language Models (LLMs) for using secure and cost-effective in-house chatbot solutions within organizations. Through the utilization of the data-privacy centered PrivateGPT architecture and the capabilities of open source LLMs, we propose a solution that addresses data security and control problems that are frequently connected with commercially available chatbot solutions. The study investigates how businesses may use these open-source tools to create unique chatbots that can be deployed inside their local infrastructure and have access restricted to only authorized individuals. The research attempts to show the feasibility and affordability of this strategy by comparing the development, deployment, and maintenance expenses of this in-house solution to choices provided by third parties. This project promotes the broader use of LLMs in the workplace landscape by offering a safe and affordable substitute for current commercial solutions, hence meeting the growing demand for responsible AI adoption inside enterprises.
Keywords: Chatbot,Large Language Model,Natural Language Processing ,AI-Based Assistant
Tools: Anaconda,Jupyter Notebook,Python,GCC (GNU Compiler Collection),Cmake,Chocolatey,llama-cpp-python,Poetry
Department: Department of Business Studies
Project Team Members
| Name |
Email |
|
Luqman Ejaz
|
luqman2020@namal.edu.pk |
|
Muhammad Danish
|
danish2020@namal.edu.pk |
|
Syed Musharaf Hussain Shah
|
musharaf2020@namal.edu.pk |
Project Poster
