SHAi, which stands for Sustainable Hydroponic AI, is a groundbreaking project to revolutionise hydroponic farming practices. Focused on sustainability, this domain-specific Large Language Model (LLM) is designed to cater to the diverse needs of hydroponic enthusiasts while aligning with eco-friendly agricultural methods. Key Features: Comprehensive Guidance: SHAi provides comprehensive guidance for hydroponic farming, covering everything from basic principles to advanced industrial applications. Sustainability Integration: Aligned with sustainable agricultural methods, SHAi promotes eco-conscious practices, contributing to a more environmentally friendly approach in hydroponics. Domain-Specific Expertise: Tailored specifically for hydroponics, SHAi ensures that users receive specialized and relevant information, fostering success in their farming endeavors. User-Friendly Interface: The SHAi interface is designed for accessibility, offering an intuitive platform for users at all skill levels, from beginners to experts. The SHAi project leverages the LlamaIndex framework to streamline the Retrieval Augmented Generation (RAG) process, enhancing the querying of the Zilliz vector database. With the assistance of LlamaIndex, context and prompts are efficiently extracted, enriching the information passed to Vertex AI's PaLM (Parameterized Language Model) for generating more prominent and relevant responses. This synergy between LlamaIndex and RAG ensures that the Language Model eliminates LLM hallucinations and accommodates a diverse audience, offering tailored guidance across proficiency levels in both home-based and industrial-grade hydroponic farming. The entire system is seamlessly integrated into a user-friendly Streamlit app, further enhancing accessibility and user interaction.Category tags:
"Great use of all of the technologies and interesting use case. Would have liked to see a broader set of evaluations used that validate the performance of your application to some specific business requirements outside of the hallucination/rag triad metrics."