Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> How would you sell what I've built?

It's interesting. You have built something tightly coupled ("like a classic domain controller") but then it is interacting with inspecific, totally decoupled stuff ("(p2p vpn), L4 (mTLS), and L7 (OIDC)").

"Tightly coupled for me, but not for thee" - why would someone who has adopted a decoupled application infrastructure decide that their domain controller should be coupled? I feel like people want one or the other in totality, they are either completely a Windows shop, or they are completely using bits and pieces of everything from everywhere. Everyone in between is ultimately migrating to one end or the other.

I can't speak for how to sell something I've never used. But I know Okta is very popular, and I encounter many IT people in many tech forums basically describe a feature of Okta. That's a huge scope. But that's a company that has tackled the dichotomy of coupled versus decoupled solutions, by simply providing everything. Is there a little bit of a chance that a single person can make something competitive with Okta? Yes!



> why would someone who has adopted a decoupled application infrastructure decide that their domain controller should be coupled?

You'd really only want the current appliance if you don't have the in house staff to assemble/amalgamate an equivalent setup.

My pitch would be that tight coupling enables major security benefits in the implementation:

* rapidly propagate policy updates (eliminate race conditions caused by changes slowly propagating across SaaS vendors.) * simpler modules with fewer features, less attack surface * rich context in logs (even spoofed packets have a cryptographically verified source) * coherent security controls

> something competitive with Okta

I could imagine building the product into "the Okta of on-prem".


> "the Okta of on-prem"

OK, now that sounds like a real product with a market, unlike your original post. I don't know if or how you can bootstrap to there (I assume Okta took a bunch of money so that they could blitz out to integrating with everything and didn't need to make sales when they couldn't integrate with everything - maybe building the integrations with a customer's systems as you acquire that customer?) - and I don't know how many people have non-legacy prem systems that they're willing to spend money on. But this is the pitch.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: