-
-
screenshot of AishBot Demo
Inspiration
The inspiration behind AishBot came from the desire to create a highly interactive and responsive AI assistant capable of understanding and generating human language with a focus on academic support and course suggestions. We wanted to leverage the advancements in AI to provide users with a platform that could offer immediate, accurate, and insightful responses to a wide range of academic queries. Additionally, recognizing the growing demand for professional courses, we aimed to integrate course suggestions from IQinternz, thus providing users with a valuable resource for their learning and professional development.
What it does
AishBot serves as a multimodal AI language model, hosted on a Streamlit application, that interacts with users in real-time. It is designed to comprehend user queries and generate detailed responses across various academic subjects and recommend courses offered by IQinternz, such as Web Development, Python, Java, Cyber Security, RPA, AI, Data Science, and Pharmacovigilance. The integration with Google's generative AI allows AishBot to provide tailored advice and support, enhancing the user's learning experience and career planning.
How we built it
We constructed AishBot by meticulously fine-tuning Google's generative AI using a dataset of over 500 structured prompts, each carefully crafted to encompass a broad range of academic inquiries and course recommendations. This fine-tuning process was crucial in tailoring the AI's responses to be both accurate and contextually relevant to the user's needs. After configuring the AI with a specific API key and establishing an initial prompt to define AishBot's domain of expertise, we integrated this enhanced model within a Streamlit web application. The core functionality of AishBot is encapsulated within the ai function. This function not only processes user inputs but also interacts with the fine-tuned generative AI model to produce responses. It meticulously logs the conversation history, ensuring that each interaction is context-aware. Such an architecture facilitates a dynamic and intelligent dialogue between AishBot and its users, making the Streamlit application a powerful platform for academic support and personalized course guidance.
Challenges we ran into
One of the main challenges was ensuring that AishBot could maintain context across conversations to provide coherent and relevant responses. Integrating Google's generative AI models and maintaining state within a Streamlit app required careful management of session states and conversation history. Additionally, fine-tuning the AI model to accurately understand and respond to a wide range of academic queries while also suggesting specific courses presented a significant challenge.
Accomplishments that we're proud of
We are particularly proud of creating an AI assistant that can accurately understand and respond to a diverse set of academic queries. The ability of AishBot to suggest relevant courses from IQinternz based on the user's interests or queries represents a significant achievement in providing personalized education and career advice. Additionally, successfully integrating this AI within a user-friendly Streamlit application, ensuring a seamless interaction experience, is another accomplishment we cherish.
What we learned
Throughout the development of AishBot, we learned about the complexities of building conversational AI systems, especially the importance of context management for maintaining coherent dialogues. We gained insights into leveraging Google's generative AI for natural language understanding and generation, and the intricacies of integrating such technologies within a web application using Streamlit. Furthermore, we deepened our understanding of user experience design, aiming to create an intuitive and engaging platform.
What's next for AishBot
The future of AishBot looks promising with plans for further enhancements and new features. We aim to expand the knowledge base and capabilities of AishBot to cover more academic subjects and provide even more personalized course suggestions. Improving the AI's understanding of user context and previous interactions for more coherent and relevant responses is also a priority. Additionally, exploring the integration of multimedia content and interactive learning resources could significantly enhance the user experience. Ultimately, we envision AishBot evolving into a comprehensive educational assistant, supporting users in their learning journeys and career development.
