LocalExtract: Self-Hosted AI Video & Audio Summarization Engine

TLDR: 1. Site performance, inspect it carefully and make sure it's usable across devices. Im not coming back to something that's unusable 2. Competitors, consider what you're up against with respect to pricing. Be realistic about what it costs you to run. Limited credit model at 11.99 versus the leading model providers at $20/month and far more use case span The site looks good man but I'm literally just trying to check out the different plans or navigate from the homepage and the site is unusable. I thought it was just the Reddit browser but went to chrome (mobile) and the homepage takes at least 2 minutes to render content, I can't get the hamburger menu to open at all. Just an assumption based off of this: how do you gather feedback? What about measuring latency for processing times for your users? Any measurement of how users are actually using your app? If I was a user who had to come back to these wait times every time when I could just pop a link into any AI tool for the same use I would probably be very hard to retain. I would start by running some analysis of your integration latency. I'm not saying that regular models replace what you do, but what you should be selling here is unique, purpose built processing of videos/podcasts/etc. I can't speak to your processing times but if your website is any glimpse into it I would start there. Depending on your tech stack this could be worth running for free still. If youre running your own infra this is probably cheap to run (if no one is making API calls, nothing to process, no real cost) but if you subscribed to a bunch of managed providers then cost might be eating you alive. How are your margins? I finally got to pricing by just entering /pricing but couldn't get the menu to take me there within a reasonable time. Checking that out I'm concerned about what the "credit" model actually buys me. I see the FAQ lays it out. Just consider what you're up against: max plan at 11.99 with a limited credit model or I just pay for a model at $20/month that does more than just this use case and create a skill.

I feel like with a good prompt, couldn't I just do what your service does with any AI agent?

TLDR: 1. Site performance, inspect it carefully and make sure it's usable across devices. Im not coming back to something that's unusable 2. Competitors, consider what you're up against with respect to pricing. Be realistic about what it costs you to run. Limited credit model at 11.99 versus the leading model providers at $20/month and far more use case span The site looks good man but I'm literally just trying to check out the different plans or navigate from the homepage and the site is unusable. I thought it was just the Reddit browser but went to chrome (mobile) and the homepage takes at least 2 minutes to render content, I can't get the hamburger menu to open at all. Just an assumption based off of this: how do you gather feedback? What about measuring latency for processing times for your users? Any measurement of how users are actually using your app? If I was a user who had to come back to these wait times every time when I could just pop a link into any AI tool for the same use I would probably be very hard to retain. I would start by running some analysis of your integration latency. I'm not saying that regular models replace what you do, but what you should be selling here is unique, purpose built processing of videos/podcasts/etc. I can't speak to your processing times but if your website is any glimpse into it I would start there. Depending on your tech stack this could be worth running for free still. If youre running your own infra this is probably cheap to run (if no one is making API calls, nothing to process, no real cost) but if you subscribed to a bunch of managed providers then cost might be eating you alive. How are your margins? I finally got to pricing by just entering /pricing but couldn't get the menu to take me there within a reasonable time. Checking that out I'm concerned about what the "credit" model actually buys me. I see the FAQ lays it out. Just consider what you're up against: max plan at 11.99 with a limited credit model or I just pay for a model at $20/month that does more than just this use case and create a skill.