Happy Tuesday!

This week, I learned how to create my own custom AI app.

What I Created

I developed a unique AI app that generates character summaries of characters from Grey's Anatomy. This app, built with the Python library LangChain, stands out for its ability to understand and communicate using human language. It achieves this by connecting to powerful language models, which are systems that can process and generate human-like text.

The large language model (LLM) I leveraged for this project was OpenAI's GPT-3.5 Turbo. Many of you are probably familiar with this model, as it's one of the models used by the popular chatbot ChatGPT.

The Process

With my intermediate knowledge of Python, I found it surprisingly easy to learn LangChain. The documentation on their website and a few YouTube videos were all I needed. Another tool I picked up for this project was Streamlit. This open-source Python library made creating interactive web applications a breeze, even for someone like me with minimal coding experience.

Streamlit simplified the process of creating a front-end user interface where users can interact with my application. To create my app, I used my OpenAI API Key to connect to OpenAI's GPT-3.5 Turbo model. Then, using LangChain, I created a prompt template that would accept user input (a character name) and return a summary of that character using natural language. Finally, I used the Streamlit library to create a user interface.

The Results

Demo

Here is a brief demo video of the app I created.

Link

Want to try out the app for yourself? Here is the link!

Tutorial

Interested in learning how to create an app like this. Here is my first Youtube video; a tutorial on how to create an AI application with LangChain, OpenAI and Streamlit.

Conclusion

Now that I've built a solid foundation with LangChain, I'm eager to explore its potential further. I plan to create more AI applications with practical applications in my daily life. The possibilities are endless, and I can't wait to see what I'll make next!

Tuesday Top Three

Youtube Video

buildspace - "branding 101: let's get personal."

Great lecture about personal branding, where I learned the three most important things about creating a personal brand

• Create a great bio

• Anticipate what people ask when they read your bio

• keep your feed consistent with your bio

Youtube Video

The Wall Street Journal - "OpenAI's Sora Made Me Crazy AI Videos--Then the CTO Answered (Most of) My Question"

Interesting interview with the OpenAI CTO discussing Sora, the company's new text-to-video generative model

Article

Towards Data Science - Designing RAGs

Great overview of what it takes to design a great RAG (Retrieval-Augmented Generation) app

• Choosing the right components

• Designing the indexing method

• Choosing a synthesis model