This post is the first part in a series whereby I document my progress on my side project.
Large Language Models (LLMs) have been all the rage, especially since the release of OpenAI's ChatGPT: a web-based chat application that utilises Generative Pre-trained Transformer (GPT) models to provide one-on-one conversational prowess, including understanding context, generating coherent responses, and adapting to various conversational topics.
However, tools like ChatGPT currently offer one-on-one chats and often come with a price tag. This limits who can use them and how. Imagine having a chatbot in group chats on apps like Telegram or WhatsApp. It would mean more people could benefit from just one subscription. Plus, a chatbot in a group would need to respond to different topics and voices, making it even smarter and more useful. This isn't just about making chatbots available to more people; it's about making them fit better into how we talk online today, and better integrating them into our workflows.
After laying down the broader motivations for such an application, I'd like to share some personal driving forces behind embarking on this side project of building a Telegram bot powered by GPT.
Gaining System Design Experience: As I navigate my final year in university, I've recognized a gap in hands-on system design experience in many academic curriculums. Taking on this project isn't just about addressing an industry need, but also about furnishing myself with a practical understanding of building scalable systems. It's one thing to read about system design, but entirely another to implement it from scratch.
Exploring Diverse Technologies: The realm of technology is vast, and every project can be approached with myriad tools and methodologies. My aim is to craft an affordable and resource-efficient system, leaning heavily on open-source solutions. This doesn't just democratize access to models like GPT, but also grants me an invaluable learning opportunity, enabling me to familiarize myself with a range of technologies.
Diving Deeper into Natural Language Models: While the surface-level applications of models like GPT are fascinating, the underlying mechanics and intricacies intrigue me even more. This project serves as a deep dive into understanding natural language models, their workings, and most importantly, tailoring them for real-world applications.
In essence, this isn't just a side project. It's a personal journey of exploration, learning, and growth, bridging the gap between academic knowledge and practical application. So, come join me as I build a scalable, functional GPT-based chat bot that can participate and collaborate with you in your group conversations: BotTalk! This series will serve as a dev-journal as I develop my application, as well as keep me accountable for the progress that I make.
Transitioning from motivation to action, I've envisioned this series of articles to closely align with the stages of the Software Development Life Cycle (SDLC). Each instalment will not only mark a developmental milestone for BotTalk but also offer insights into the intricacies of each phase. With our next piece, we will be immersing ourselves in "Requirement Gathering", laying out foundational needs, and sketching the personas BotTalk seeks to serve. As we journey forward, upcoming articles will delve into system design, development, testing, and deployment. Through these chronicles, my goal is to provide a panoramic view of both the technical landscape and the myriad decisions and challenges that punctuate the creation journey. Whether you're deep into the tech realm, just dipping your toes, or simply curious about the evolution of a chatbot, I invite you to accompany me on this meticulous journey, watching an idea mature through each SDLC phase.
Thank you and see you in the next article!
Top comments (0)