Hey y'all - I've been in the lab over the past 9 months learning this AI coding space, and just yesterday I finally released my first app into the wild, and I'm looking for 10 beta testers to try it free for a month and give me some feedback.
hAIro is a virtual scribe tool that:
* organizes WA chat export .zip files into an interactive / sortable / filterable chat viewer
* translates / transcribes audio/video (using OpenAI Whisper and LLM diarization)
* Pastes transcript into the WA chatlog for voice memos and shared media (or just directly transcribes/translates individual A/V files
* Uses ffmpeg to convert any A/V file type into any other A/V file type
* Uses nconvert to reduce file size or enhance image files
It's currently wired up with all my personal API secrets on a VPS, so I'm limiting my beta testing group to 10 users to monitor costs, but all testing will be free.
It uses OpenAI API to access Whisper. I use OpenRouterAI to access LLM for processing the transcript to diarize by speaker. ffmpeg and nconvert are both direct access with no cost for now (nconvert requires a license fee if I commercialize, but for testing, it's all good under their personal use policy).
This project is something I built as an in-between solution, as I've been working with another dev to build a SaaS version of the WA chat organizer function, that is live-linked to a WA account, sync's in real-time and creates an interactive chat interface that is functional...
It's taken a bit longer than I hoped to get the chat tool ready to go, and I needed a local solution, so first I just built it on a local dev server, and it was really really useful.
So I decided to deploy to my VPS, and create a user account set up for multiple users.
If the testing goes well, and folks see this has some value, I'm going to put a low price one-time fee option for folks that want to bring their own secrets, or a monthly subscription cost, for users that want access but will use the secrets I pay for usage on, with usage limits that I'll add in over time.
I've also thought about putting the repo in a zip on gumroad and just selling access to a sanitized codebase for folks that may want to ideate and take it on their own journey...
A few sample use cases that I love using this for...
* Yesterday an architect I'm collaborating with on a large land dev project, sent me like 40 messages back to back, that all had massive files attached, PDFs, images, xls cost worksheet, etc... to use a portfolio content for a project proposal we're putting together. Rather than spend 45 minutes in whatsapp, downloading, and all that - I just exported the zip, processed it with hAIro, and had everything at my fingertips within a few minutes.
* There's a Bali community WA group that many of us are in that Salman created, and there are tons of amazing referrals, links, and other Bali life tips that I often remember they're in there, but I can never find the ones I'm looking for, so I asked Salman to remove chat privacy so I could export, and now I have this epic interactive chat-viewer that I can easily find all the contacts, all the links, and all the files / images shared in that group chat.
* When folks share vcf contacts, the VCF doesn't always transfer correctly, and doesn't always read-out all the information - so I have the hAIro tool set up to scrape the VCF, and print the full info as text in the chat log.
* WA doesn't allow mp4 video files to be shared, but you can share M4a's, so I just run the mp4's that I create in my video editor through the hAIro converter tool, and within seconds I have the M4a that can be shared.
* Some audio translation/transcription tools don't allow .opus files from WA to be uploaded, but hAIro can convert opus to MP3 in seconds. hAIro also reads opus directly, so I don't need a 3rd party tool anymore for any of my transcription or translation needs.
* I'm doing alot of web design lately, and there are different upload size limits for various features and sites, so I'm often time having to reduce image file sizes, which in the past I would just web search a third party tool, (I eventually downloaded XnConvert, but it's inefficient app...) - so now with the nConvert wired in Studio tool, within seconds I can upload massive image files, like the 128mb + TIF files that I have for hi-res print products that I sell, and I can reduce and convert to JPG in seconds using the Studio app to reduce to 10MB, and the Converter to change from TIF to JPG for uploading.
There are more ways I use it, but overall, it's already saved me a $99/yr Otter subscription, and probably 100 hours of time in the last 3 months that I've had it on my local machine... and since I've now got it deployed to a VPS at my custom domain, I don't have to waste the 10 minutes launching the app on a local dev server now everytime I need it...
Anyway - would love to hear your thoughts - if you want to help with the beta test, just create an account. I have to approve accounts manually so message me if you create an account so I can do that - the 1st 10 to sign up will get 1 month free testing access, and then I'll decide where to go from there.