Hello, I am a new developer and new to this forum. I am trying to automate system processes at my organization. I need to do image captioning and optical characterization recognition and summarize comments of various products and establishment. The aforementioned data coupled with basic details will be generated into a csv file. We then would like to place that csv file in a RAG to enable employees to ask questions of the dataset. From the OpenAI documentation this seems to be about a combination of OpenAI API and ChatGPT Plus capabilities. Several posts have stated that the OpenAI API gets expensive quickly and I am not sure it would be worth both resources. I estimate we would process about 1500 pictures, 5000 - 10000 reviews, and the csv would have about 150 entries with 10 data fields. Welcome an insight from experienced users.
Welcome to the community!
ChatGPT is more of a consumer product. At the scale you’re talking, you’ll likely want to develop your own system using the API.
The quickstart page is a good place to start…
Your requirements sound roughly like a hundret dollar max in API cost.
You might need another hundret to figure out how it works.
1 Like