Works

Portfolio

AI Agent With Python & LlamaIndex & OpenAI API

Portfolio

Explain

Nowadays, there are many projects developed together with AI. I saw that AI Agents were becoming widespread, especially in startups, and I developed an AI Agent that analyzes pdf and csv files and messages and replies with you. I made this AI Agent using python, OpenAI API, LlamaIndex, Pandas.

Before developing the project, I conducted various research in this field. I read the OpenAI API documentation and LlamaIndex documentation. Then, I downloaded the functions I needed for the project and downloaded the datasets from Kaggle. I made csv files readable with Pandas. Thanks to LlamaIndex, I created 2 engines. One of them is for PDF analysis and the other is for csv analysis. In fact, we can say that it is not analysis, but engines that extract relevant information from relevant files. Additionally, at my request, I had him write down the results he gave me in a file. I created an engine for this. I have taken prompt engineering courses for a long time. I used the knowledge I gained in this project. Again, thanks to the PromptTemplate offered by Llamaindex, I created prompts to perform these operations.

html5 bootstrap template by colorlib.com

After running the file, I can now enter inputs from the command line. It performs the relevant work with the Openai API according to the prompts I mentioned before.

Go to github