import SocialEmbed from "../../../../components/SocialEmbed.astro";
# Interview with Rody Davis - Firebase
<SocialEmbed platform="youtube" id="D2q_y1GFAIo" />
## Description
Rody Davis is a Developer Advocate for Firebase Studio at Google. Working closely with Flutter, Angular, Firebase, Go and Gemini in Google Cloud. In my time at Google I have also supported Material Design, Lit (previously Polymer) and Google Developers (Solutions). I have a background in systems administration, audio engineering, music and electrical engineering. I am very passionate about open source and creative applications of software. Currently I am working with teams to take advantage of AI in software development and integration into apps. My mission is to make the intersection of Google products work better together.
## Transcript
**0:00** · And we are live here at Narala and we have Rody Davies with us. Hi Rody.
**0:04** · Hi. How's it going?
**0:05** · Rody is developer relations uh engineer for Google at Firebase. So we're going to talk a little bit about Firebase and your experience here in the lab.
**0:14** · So yeah, of course.
**0:16** · So in simple terms like what's Firebase and why does it help people to do things? Yeah, one of the cool things about Firebase is we are in Google Cloud trying to make it easier for you to connect to all sorts of APIs that can be traditionally hard if you're not an enterprise developer. So, if you're just wanting a mobile application and you don't want to have a backend, but we make it as easy as possible to start with a database authentication, sending out push notifications. We've launched a lot of new features like um protection around if you have adversarial people trying to get into your app.
**0:48** · We also have a way to connect with AI. So Firebase AI logic is a way to connect to AI securely without having to worry if you're going to send your API key to the client.
**0:59** · I mean, and it's just for developers, Firebase or everybody can use the the tools and services.
**1:05** · So Firebase is traditionally targeted at developers, but with the launch of Firebase Studio, we really started to broaden that definition to just builders. We see a lot of people coming to Firebase Studio building an app for the first time or coming back to development after 101 15 years because now they can just give it an idea of what they've been wanting to build maybe a game for their spouse or what's really cool is we see a lot of data scientists upload uh a lot of data and say like I just want to visualize this. I want to be able to solve this custom workflow.
**1:36** · And so we've seen some really cool use cases we didn't expect when we originally started building it. I was going to ask that like it's like you share internally like hey we saw this we saw that like this is amazing we didn't thought about it like it's usual.
**1:47** · Oh of course. Yeah we we really like being able to share. We even have on Fridays a customer demos. Well some of the best ones from the community will come and present to the team too.
**1:56** · Oh that's I mean awesome seeing other people do things with tools then it will bring innovation all over the place.
**2:02** · That's that's awesome.
**2:03** · And do you have businesses using Firebase too? because like developers will be maybe the gate to the to the tool and then it grows into the companies.
**2:10** · So with Firebase, we already have millions of businesses that are building on top of Firebase. It's already really popular. U with Firebase Studio, we've already seen just in the short time since we launched since Cloud Next this year. We've already seen multiple people say that they've not only built an app with Firebase Studio, but they've launched it to production and are already having customers.
**2:29** · I mean that's that's really that really great. And now let's go a little bit into your personal experience with Firebase. We were talking a little bit off camera like what's your favorite use cases? I mean what's what do you love most to do with this and now with AI addon, right? Like what do you prefer to do with AI on Fire?
**2:45** · So AI obviously is a fastmoving pace and I think that there's some really cool things going on in terms of innovation, how we're using context effectively. One of the things that I have been really excited about with Google and Gemini is things like Gemini deep research notebook LM where you can take all the context and create your own knowledge database. But then I like to take it one step further and create some really cool custom workflows with Gemini CLI. So I can take my documents, use a research plan, and then create new context to send to the model.
**3:15** · So I'm basically compressing exactly what I need at any one time while being able to pull from all my resources. So like at a company you may have a PRD and a description of the designs and the different features that you want to implement. But so it's important to put that into a text format that the model can understand. But one thing that I think uh people may not realize is subtraction is actually more powerful than addition when it comes to AI. If you reduce a lot of the context and are only giving it the relevant bits, you're going to have a way better result.
**3:46** · Even with Gemini having a 1 million context window, we found that effectively if you're using about 20% um you can get a lot more performance out of that that context window, it's really good for searching long PDFs and being able to do some really cool context, but to get that high performance, you can also take advantage of the implicit caching and things like that too.
**4:06** · That's great. And and if we go just to the new features that we have on Firebase, what would be your maybe top one or top two new features in the last time? So, one of the cool things that we launched really recently is Firebase AI logic, which allows you to proxy all your requests to our back-end servers.
**4:23** · Um, traditionally, you've we've you've been able to like build with AI using a lot of the common SDKs, but people are always worried about sending their API key to the client. Well, Firebase Logic will actually proxy it, so you never have to worry about that. But you can still run inference on Google Cloud through Vertex or the free tier in Geminina API. So we've seen some really cool use cases. Not only that, we partner with Chrome. So now on certain devices, it'll even run the ondevice model of Gemma or Gemini Nano actually.
**4:53** · So you can completely run offline and then fall back to that cloud hosted model. We're trying to bring that to other mobile devices too, but it has been such a cool use case that like you just write your API and then your your API costs go down if they they have a supported model.
**5:08** · That's awesome. Yeah, a lot of things going on and it's really exciting to be right now with AI and everything going back to the device or the browser. So just as a final question like what would you recommend for someone that wants to begin in Firebase like do they have communities resources online? Just a brief comment on that.
**5:24** · Yeah, of course. So Firebase is a massive community with a lot of people all over the world. Um if you're in an area that we um have supported which is almost all places you can look for what are called Google developer experts and we have communities here that you can get started with meetups and they'll connect you with all kind of uh resources. We also work with those people so we can uh make sure that we have content that's scalable and then of course uh there's Googlers you can find the events we have a lot of content online code labs pathways you name it.
**5:54** · That's awesome. Thank you for the time here, Rory.