Software Architecture for social media scanning project

Hi, we've narrowed our "scan social media posts, looking for keywords" project down to 2 people/agencies. (you and 1 other) This is our 2nd attempt to hire for this project. The 1st time, the developer was over his head. This time, we need to make sure you have a clear, detailed plan of how you're going to build it. And we'll pay $25 USD for your time and expertise. Client, server, input, output, db, caching, storage, APIs, etc. The more detail, the better. The other thing to keep in mind is that we plan to add OCR to v2, if this v1 is successful. We're evaluating AWS Textract, Dynamsoft, a9t9, and Cloudmersive. This will let us scan images/memes/video in v2, as well as the text in a social post in v1. Make sure you think about adding OCR in v2, because we'd like the same person to help us with that, as well. And if you don't think you can handle it, please let us know. Take a look at the architecture examples below and let me know if you have any questions