GRVYDEV

Thanks! I actually prefer a separate window personally since I find myself always wanting to quickly be able to reference this app. With it being its own application I can quickly use spotlight search to focus the window instead of needing to dig through my browser tabs to find it

Thanks again for this suggestion. I just went and added these features

This isn’t the 90s anymore. Using the systems web view is, in fact, native by definition.

I was using obsidian before this. There were two main frustrations for me with obsidian. First, opening a markdown file that did not live in a vault wasn’t possible. Second there was not a CLI that made it easy to quickly open files

I live in my terminal but for some reason any TUI markdown viewers just don’t do it for me

Haha awesome. Was the lowest hanging fruit for a name

You should ask them to make it work on windows or Linux once they’re done on Mac :) would be a good lesson for them

Voiden looks awesome I’ll have to check it out

I’m a neovim user so I’ve never used the vscode markdown viewer. One of the things I want to add is the git and live update features though. I think those would be a game changer

These are great suggestions I’ll get these added tomorrow!

Awesome that’s a super interesting approach

Oh nice! I’ve never checked out Typora before I’ll take a look as well

Dropped you an email!

You can aggregate data from multiple different sites with the hope that it gives you a true picture of what the company does

Thank you for this thoughtful response. We have the same mindset. This is exactly what we are building. I’d love to chat with you more on this.

This is a great use-case!

Thank you for the feedback. Apologies for the crappy response. This is still super nascent technology and I’m still working on improving the search process. In my eyes the real value here lies in data enrichment from other sources allowing you to get a really deep understanding of a company. I am actively building out that functionality now.

Thank you for the feedback! I'll get to work on an About page.

What we are doing is automating bespoke account qualification and research. Say, for example, you are a company that sells software to chemical synthesis labs. Your ideal customer synthesizes more than 5 chemicals. They need to synthesize them in large batches and they need to have more than 10 chemists on staff.

To go out and find those ideal customers, right now, you would need to spend hours upon hours searching and manually qualifying each account. We are going to automate that. By providing more accurate leads we will reduce the amount of time that companies need to spend on prospecting while also increasing the open rate.

Id love to add hotword detection! Or, even better, you (or someone in the community) could add it :)

Hey! First of all thank you for this really detailed response! I am very new to the voice space and definitely have a TON to learn. I'd love to connect and chat with you sometime :)

I totally agree with you about latency. It is very very important for use cases such as a voice assistant. I also think there are use cases in which latency doesn't matter that much. One thing I think I may have understated about S.A.T.U.R.D.A.Y. is the fact that, at it's core, it is simply an abstraction layer over vocal computing workloads. This means it is 100% inference implementation agnostic. Yes, for my demo I am using whisper.cpp however there is an implementation that also uses faster-whisper.

I also want to call out that I have spent very little time optimizing and reducing the latency in the demo. Furthermore, when I recorded it I was on incredibly shoddy WiFi in northern Scotland and since the demo still depends on OpenAI a large chunk of the latency was introduced by the text-to-text inference. That being said there is still a ton of areas where the latency in the current demo could be reduced probably to the neighborhood of 1s - 1.5s. This will get better in the future :)

I want to touch on something else you mentioned. GPUs. I intentionally tried to avoid using any GPU acceleration with this demo. Yes, it would make it faster BUT I think a large part of making this kind of technology ubiquitous is making it accessible to as many clients as possible. I wanted to see how far you can get with just CPU.

In regards to your comments about NLU/NLP I have not dug into using them in place of LLMs but this seems like an area in which I need to do more research! I am very far from an AI expert :) I have a bunch of ideas for different ways to build the "brains" of this system. I simply have just not had time to explore them yet. The nice part about this project and demo is that it doesn't matter if you are using an LLM or an NLU/NLP model, either will plug in seamlessly.

Thank you again for your response and all of this information! I look forward to hopefully chatting with you more!

More-so an oversight :D

It's made with https://excalidraw.com/

Thank you for the very kind words :)

Yeah adding memory is absolutely on the road map! For now everything is quite simple to allow for us to build over time but I plan to put a lot of work into the text to text engine in the coming months

If you really think about it every future is near term with a shelf life :)

Yeah! There is a YouTube video linked in the repo. I’ll edit the description to add it to the HN post as well :)

This is awesome. So happy to see this technology maturing. Great job team

I went with FTL instead of RTMP for the sake of latency. Also FTL gives me a stream of RTP packets which can go directly into WebRTC meaning I have to do 0 processing of the packets where as with RTMP I would have to convert them into RTP packets. Also SRT is interesting however it is wildly complicated and does not use RTP meaning I would have to figure out how it works and then convert whatever it gives me into RTP packets for WebRTC

Yup I have spoken to them about that as well however that move wont be for a while

Yup that was the point :) In the future configuration of the server will become easier as well!