𝗛𝗼𝘄 𝗧𝗼 𝗕𝘂𝗶𝗹𝗱 𝗔𝗻 '𝗔𝗜 𝗳𝗼𝗿 𝗩𝗖' 𝗔𝗽𝗽 𝗜𝗻 𝗠𝗶𝗻𝘂𝘁𝗲𝘀 I've worked on 40+ editions of AI for VC over the past year. I decided I wanted to turn these tutorials into their own web apps. Here's how I did the first one 👇 𝟭. 𝗙𝗶𝗻𝗱 𝗔 𝗨𝘀𝗲 𝗖𝗮𝘀𝗲 One of my favorite newsletters was this one on turning a market map into a spreadsheet: https://lnkd.in/e2dHQJqC Only one problem. It was a multi step procedure... My standards are getting higher. Let's turn those 3 or 4 steps into 1. 𝟮. 𝗔𝗹𝘄𝗮𝘆𝘀 𝗦𝘁𝗮𝗿𝘁𝘀 𝗪𝗶𝘁𝗵 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 💬 Here’s the first prompt I ever sent on this topic: “I want to build a web app for turning an image of a startup market map into a CSV file.” (See page 2) Now I have a bit of a framework for what this is going to look like. But I got greedy. I wanted ChatGPT to do everything. After a little more back-and-forth with ChatGPT, I started to get a good idea of the tech stack I would need. 𝟯. 𝗧𝗵𝗲 𝟯 𝗧𝗼𝗼𝗹𝘀 𝗙𝗼𝗿 𝗧𝗵𝗲 𝗝𝗼𝗯 🛠️ ChatGPT: We’ll use ChatGPT to help us write the code for our new web app. We also need an API key to send requests to OpenAI. You can get yours at https://lnkd.in/e8XRZtW9 PyCharm: PyCharm is a development environment for Python projects (the language we’ll use). Don’t worry if you don’t have much Python knowledge. You can download the IDE here: https://lnkd.in/eAWy-nHj Streamlit: Streamlit is a framework for deploying, managing, and sharing your apps with the world. This is where our final application will be hosted. Get a free account here: https://streamlit.io/ 𝟰. 𝗧𝗶𝗺𝗲 𝗙𝗼𝗿 𝘁𝗵𝗲 𝗛𝗲𝗮𝘃𝘆 𝗟𝗶𝗳𝘁𝗶𝗻𝗴 🏋️♀️ After getting familiar with our tech stack, our starter prompt evolved: “I want to build a Streamlit app using a Python script for turning an image of a startup market map into a CSV file. Please write the necessary code for me to do so. I would like to use OpenAI’s model 4o.” (See page 3) Naturally, I encountered a few bugs and glitches along the way, but it didn’t take long to resolve and update the code to get this program working. I copied the code and pasted it into a new Python file in PyCharm (See page 4) Before I ran the code, I needed to make sure all of the libraries were installed properly. I went to the terminal (see page 5) Then typed: “pip install streamlit requests pandas pillow” That did the trick. Now it’s time to run my application. 𝟱. 𝗧𝗵𝗲 𝗙𝗶𝗻𝗮𝗹 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 📲 I used the terminal to run the program with this line: “streamlit run marketmap.py” A new window opened up in my browser (see page 6 for the result) Here’s a quick demo I put together to show how to run the program from inside PyCharm: https://lnkd.in/eveRFFhV LMK if you'd be interested in the final code!
Hugo Rauch
6mo
This is super cool!
Vice President - Data Products @ PNC
6moHey you forgot to close out of chessbot2 champ