I secured the domain and used AI agents to build out a functional proof-of-concept for the client. To handle infrastructure, I routed the application logic through Mailgun. Internally, the prototype is fully operational — emails successfully send and receive from @w.xyz handles. W combines a highly optimized interface with built-in AI capabilities like Spam Scanning, Smart Reply and Text Summaries.
I currently have a 30+ organic waitlist as I finish the final touches and prepare to move away from the prototype stack.
My questions for the community:
Concept Validation: Given how crowded the email space is, do you see a viable market for an AI-native premium webmail service, or do you believe the switching cost from Gmail/Outlook too high for users?
Scaling Roadmap: As a non-technical founder moving away from an AI-generated prototype stack, what are the most critical architectural baselines I must prioritize right now to ensure the system scales smoothly?
P.S. If you're interested in building a new kind of email platform from the ground up, drop a comment below.
"W combines a highly optimized interface with built-in AI capabilities like Spam Scanning, Smart Reply and Text Summaries." - add that to the homepage. It's entirely unclear to the reader.
Personally I wouldn't need any of the features, the big players already have them.
Website is hard to read. The font size, the colors. What's the fascination with dark gray text on black background?
Short domain name is only as helpful as I don't need to spell it out over the phone. hey-dot-com vs double-u-dot-ex-why-zed.
https://en.wikipedia.org/wiki/.xyz "It was estimated in 2022 that 6.62% of extant domains in the .xyz TLD were malicious. Due to these instances of malware, scams and phishing, multiple anti-malware vendors have blacklisted many or all .xyz domains."
Using Mailgun is too risky. You're responsible for your user actions and they can suspend the service when they see over 0.08% spam reports. One user doing a spam campaign can risk the whole project.
However, for the most part, an email domain's reputation relies on its specific sending habits and configuration, not just the extension it's on.
I doubt anyone is actually going to spell out w.xyz by pronouncing the letters individually, it's just 'w dot x-y-z'.
I will certainly work hard to ensure we stay firmly within the 93.38% of .xyz domains that are legitimate and secure.
As for the infrastructure, Mailgun was selected strictly to deploy a rapid, working prototype. It was never intended to be the final production software as we scale.
I could be wrong but I think very few will be enticed by the domain alone or at all.
Personally if I was building an email platform, I wouldn't rely on a third party email API as its part of your core thing, I'd run my own email stack, would give you so much more control / flexibility and be cost effective.
I'm pretty happy with fastmail for my email and use mailpace for transaction stuff from apps, although I did run my own mail servers for about 10 years.
The domain isn't the whole product, but for our target niche, a short, memorable handle is a highly valued status asset. We already have 30+ organic people on our waitlist who reserved their handles. As for the infrastructure, Mailgun is only being used for a working prototype. Using their webhooks was simply the fastest way to validate the core mechanics.
Now that the prototype is working, this is the reason I'm looking for a founding engineer to take over this functional proof-of-concept and help migrate us to a custom, cost-effective, and fully controlled email stack from the ground up.