1/ The challenges of navigating the vast and complex data landscape are real. Traditional tools often struggle with data overload, lack of effective visualization, and poor customization. ChartGPT is here to solve these challenges. Let's dive deeper 👇

2/ Use ChartGPT as your personal data analyst! First, select from one of the sample datasets: - DEX transactions - NFT lending platform aggregate borrow volume - Fastest growing AI & Analytics companies Then, either choose one of the sample questions or come up with your own.

3/ Backend ChartGPT uses @LangChain as a base. This is a library that aims to assist in the development of Large Language Model (LLM) applications. LangChain introduces “agents” which have access to a suite of tools. Depending on the user input, it picks the right tool.

4/ Agent We have created an agent that transforms data into insights by offering an assistance with: - Database connectivity (in our case BigQuery, with plans to expand to other data sources) - SQL generation and data transformation using Pandas - Chart creation using Plotly

5/ LLM and Prompt Engineering We initially started with the @OpenAI gpt-3.5-turbo. After promising results, we included it in our first working prototype. The GPT-4 model, now in testing for the next release, is slower (for now), but delivers more consistent results.

6/ Frontend We used @Streamlit as a frontend, our preferred framework for rapid prototyping. It implements a chatbot interface and allows for a reasonable degree of customization for the first release of the product. Down the line, we plan to build a custom frontend solution.

7/ Are you ready to take your data analytics to the next level? Be the first to experience the power of the ChartGPT assistant! Join the waitlist 👉

8/ Product Hunt We're thrilled to share this journey of data exploration with you, but there's so much more to ChartGPT! We'd love for you to join us in refining and improving our product. Visit our Product Hunt 👉