Inspiration Lervis Enterprise was inspired by the difficulties faced by non-technical users who struggle with traditional data analysis methods. These methods often require technical expertise and can be time-consuming when done manually. The team saw an opportunity to make data analysis more accessible and efficient for non-tech users.
What It Does Lervis Enterprise transforms how users interact with data by allowing them to use natural language queries instead of manually writing SQL queries. The system offers automated SQL generation, real-time data visualization, and image generation for marketing purposes. Additionally, it incorporates Human-in-the-Loop (HITL) for report validation and integrates both structured and unstructured data sources for context-aware insights.
How We Built It The project was built using a combination of cutting-edge technologies to ensure scalability, efficiency, and seamless user experience. Key technologies include:
- Langchain, OpenAI API, Tavily AI API, Mistral AI, Groq API, SambaNova, and aiXplain for Generative AI, allowing for natural language querying and automated SQL generation.
- Plotly for real-time data visualization, enabling users to instantly analyze trends and insights.
- Streamlit as the web framework, providing a user-friendly interface, with Streamlit Cloud used for seamless deployment.
- Stable Diffusion in various versions, including v1.4, v1.5, and xl, for high-quality text-to-image generation.
- Llava and Llama 3.2 vision image generation models for creating marketing content, such as promotional images for product launches and social media. In addition to these technologies, Lervis Enterprise offers easy integration with any database and cloud platforms used by businesses, ensuring that it can be deployed seamlessly into existing systems without major changes. This makes it highly adaptable to various industries and data environments.
Challenges We Ran Into One of the primary challenges was managing token limits during the AI-powered data analysis processes. This required optimizing input and output lengths to ensure the system could handle large queries efficiently.
Accomplishments That We're Proud Of We are proud of how Lervis Enterprise has transformed data analysis by allowing non-technical users to easily engage with complex datasets using natural language. It has improved data utilization, decision-making speed, and productivity for businesses.
What We Learned Throughout this project, we learned the true power of Generative AI (GenAI) and how Retrieval-Augmented Generation (RAG) can dramatically improve data visualization and decision-making processes. It also taught us the importance of incorporating Human-in-the-Loop mechanisms to maintain accuracy and relevance.
What's Next for Lervis Enterprise The next step is to launch Lervis Enterprise as a full-fledged product. We aim to continue improving its features, such as expanding the AI’s ability to handle even more complex queries and adding new integrations to further simplify data analysis for users
Log in or sign up for Devpost to join the conversation.