> The loop runs on stateless workers in the control plane. A worker runs one iteration > of the loop, either a model call or the tool calls it produced, then releases the > agent and puts the next iteration on a queue in Postgres. ... > Also unlike other harnesses, the machines that the workers (our agent loop) run on aren't > where the agent runs its commands.
It's late but don't think I am grokking this.
Isn't running the commands the heavier part of the work? The agent loop itself is mostly waiting for a remote model and then running the commands?
Why would I need a pool of workers for the loop when the constraining resource is likely to be the machines I am running the commands against (which can't be stateless like the workers are?)
Locking in 4% AFTER TAX yield ... if you invest the $300K and it earns 4% you still can't pay the mortgage interest with those funds
I'm not a fan of OS X but seriously considering one of these just for the battery life and it-just-works portable computing.
It works great for email support, haven't tried their chat features, but is the sweet-spot for us in terms of features and complexity.
They have an API if you want to integrate, though for a web form just feeding it through email works.
Source: been there, done that ... there is no benefit to it and only pain.
Have a separate plunk-emails.com domain that is used for all emails. Or even have a bunch set up and rotate through them. This will mitigate spam impact too -- you will get spam issues even with 100% legit users
Assuming they did decide to quibble about the classification what can they actually do to an entity that has no legal existence in the US?
They do live chat too but we've never used that.
As a future product option I'd love to be able to treat these as standalone terminals, so we can offer a web based POS app (running on a different device) and the phone as a payment terminal without having to deploy our own mobile app.
Or is it more akin to the BBPOS Chipper readers and still needs a custom mobile app to interact with?
ie: can a platform built on Stripe use this for their customers without needing to supply a mobile app?
Then anything you do in OBS is feed into Zoom as a direct video feed and saves Zoom screen scraping and re-encoding the video. You can use OBS's screen capture or better yet pre-download the files and just play them back.
I think the best option might be to show something on the site when they are unsubscribed so they can easily undo that .
We provide a platform SaaS so generate emails on behalf of our customers to their customers.While our spam rate is very low, at least 2/3's of the spam reports we do get are for what appear to be totally legitimate transactional emails (eg. purchase receipts). Some could be typo's in an email address but I suspect the majority are misclicks on the end users part.Currently we unsubscribe all...
I have been unable to set up a Twilio or voip.ms number as the phone# for a Google accounts 2FA
Is it possible for a spammer to generate >$75 per 10,000 people spammed? I've no idea were the SMS spams I've got link to (not about to find out) but they are so obviously spam.
We use SMS for communicating with users and would be happy to more a lot more per text to escape the 'positive ROI for spammers' territory.
I'd be happy to do that for important emails too!
Telecom spam filtering seems to be a ridiculously primitive and wide net. I can't imagine a valid use case for dropping a text sent to a number when that number just sent you a text a few seconds before.
I don't understand why SMS spam has such a big issue with false positives compared to email spam when emails are practically free to send but SMS is much more costly.
(Yes, I know there are a lot of false positives on email too ... but we run into false positive SMS spam issues a lot even though it feels like it should be a much simpler problem to solve).
Microsoft (used to?) offer this for developer support and I remember using it maybe 15 years ago where it was a couple of hundred bucks to open a ticket but you got quick access to a real expert and good escalation.
If the issue turned out to be their problem the ticket was refunded.
For something business critical like this it is a way of signaling to the company that there is clearly somethin wrong with the automated process: a real scammer won't pony up hundreds of $ to get a review they would fail.
Who is your target customer? A small business owner without a website? In which case listing features like "Cloudfare CDN" and "DDoS protection" won't resonate with them at all.
A couple of other things that jumped out from the pricing page:
* 10,000 for submissions a month is pretty meaningless: if 10,000 people a month are interested in my business I am motivated to spend a lot more than £7.99 to capture and process them. Either make the limits really small to feel like a limit, or just say unlimited.
* support responses in 5 days ... this would be a red flag regardless of price. As a small startup the one thing you can offer is exceptional support (since you don't have many customers to offer it too!). If it takes off lower the expectations, but I'd suggest 1 day even for the lowest plan should be doable (and if not raise the price...)
A business owner doesn't really want a website builder at all, they want a website: lead more with your 'custom websites' page. If £199 is too much of a hurdle for them then realistically they aren't going to be a great customer anyway. Maybe offer a couple of business a free website to have some real world referable examples.
If you are targeting website designers (which are probably an easier to identify market) then focus more on the builder itself, maybe stressing the mobile friendly side (though I don't really see how that helps me if I am a designer ... I won't try and work from my phone).
Good luck!
Email routed through SES is getting through though.
This is going to be a mess to clean up due to automated blocking of bounced addresses
Do we stream Netflix @ 4K ... no, but it is quite workable at low res, the 1Mbps upload is the most painful portion.
My cell data is faster (but 50GB\month cap) so I will use that for large uploads to save time, and also as a backup since we get frequent fixed wireless outages (esp. in summer since any storm seems to cause problems).
I also occasionally spin up EC2 instances to remote to when moving around data to avoid dragging it down locally (usually video). Works fine ... remote desktop to a Windows instance is perfectly usable.
Zoom/Google Meet/etc work OK ... not great, but perfect usuable and a lousy connection is a good excuse to turn off video sometimes!
I hold out hope for 21st century technology some day, but wouldn't switch the trade-off of being more in-town for a faster connection.
Since sendgrid is down (https://status.sendgrid.com/) ... again, does anyone have a recommendation on a service that will abstract sending emails.We'd like a service that would present a single mail sending API, and automatically distribute outgoing emails between different providers (that we have accounts with individually).It doesn't need to send the email itself, just act as a 'load...
Looks like a mis-fire of the automated suspension system which seems to be a huge problem of these types of systems when you are on the receiving end.