Build An Audio AI App Transcripts
Chapter: Welcome to the Course
Lecture: What Could We Build?

Login or purchase this course to watch this video and the rest of the course contents.
0:00 So we're inspired to work with audio and do something amazing. What could we build? Well, possibilities are pretty endless. We can go this way or that,
0:11 but I've got a really clear idea of something super fun that we can build for this course. We're gonna build something I'm gonna call X-Ray Podcasts.
0:21 This is an app. What does it do? Well, as the name kind of indicates, it lets you look inside of a podcast. It lets you look and see more about it.
0:30 So obviously I run a couple of podcasts, but I'm also an avid podcast listener. Do you know what's really boring and not great?
0:38 Is if there's one paragraph for show notes, there's no links, not much about it. Do you wanna listen to that one or not? That particular episode?
0:47 I don't know. You don't have enough information to make that decision, do you? Well, with X-Ray Podcasts, you'll be able to get a TL DR,
0:55 a little bit of a high action point summary type of thing. What's happened throughout the show. We'll be able to get transcripts.
1:04 We'll be able to search in them, all of that kind of stuff. So that's what we're gonna build, this X-Ray Podcasts app.
1:10 And here's what it's gonna look like. Starts out, you can just enter a podcast URL here. The app is going to download, parse the podcast
1:19 all through its RSS feed, import it into the system, save in a database. And then it will start to apply its magic
1:27 through the machine learning and assembly AI platform and some search engines and those kinds of things. So here we can start by getting an app.
1:36 Then we can go into the details. For example, here's one of the recent Talk Python ones. You can see it's pulled in the show notes from the RSS feed,
1:43 but more importantly, see those purple blue buttons there? Create a transcript. Don't have one yet. And we actually supply one at Talk Python,
1:53 but many podcasts don't have transcripts, but X-Ray Podcasts, it's gonna let you create them even if the podcast person, the host didn't provide them.
2:02 Create a summary, a TLDR and important action items and even have a chat, a Q&A with the podcast content itself. That's amazing.
2:14 Click that transcript button. We'll get really cool looking transcripts like this. Press the summary button. We'll get the TLDR and other things.
2:22 Here you can see Sydney Ruckel discusses her career journey in software engineering and offers advice to inspiring programmers, et cetera, et cetera.
2:30 That's not from me when I wrote the details for the show. Sydney didn't exactly say that. Assembly AI read this and gave us this summary for us.
2:39 Super cool. And once we have all this extra information, then we can deeply search within podcasts.
2:46 If you search for a podcast today, you get show notes, you get the title, things like that. Very, very limited information.
2:55 This app that we're gonna build, this X-Ray podcast is gonna allow you to search all of that extra information
3:01 that we just talked about, surfacing through AI. So this app is gonna be super fun to build. We're not gonna build it from file, new project,
3:10 completely from scratch. That would spend a lot of time on things like web design and frameworks and database design, all those things.
3:17 We're just gonna start from an empty shell of an app and we're gonna add these features. We're gonna add search, add the transcripts
3:24 and those kinds of things. So if you're not a web developer, don't let this scare you away. We're gonna start with a pretty decently working model
3:33 that you won't have to change a whole lot for the website and then just adding these cool features, but it's still gonna be cool to see
3:40 how they surface in a real app.


Talk Python's Mastodon Michael Kennedy's Mastodon