What zero trust solution actually works for remote access to on-prem servers?

I've been wrestling with this exact problem for months. We have a few legacy on-prem servers (a file server and an old CRM database) that our remote team occasionally needs to access. The traditional VPN is clunky and feels like overkill—plus, the security team is (rightly) pushing for zero trust principles.

We've tested a couple of big-name ZTNA solutions, but they either got way too complex for our simple use case or required an agent on the server that felt invasive. I kept hearing about Twingate being "developer-friendly" and simpler, so I spun up a trial.

Here’s my quick take after a month of testing:

* **Setup was shockingly fast.** The connector on our on-prem network took about 15 minutes. No need to mess with our existing firewall rules, which was a huge win.
* **The access model just makes sense.** You define resources (like `legacy-file-server:445`) and assign them to user groups. No network-wide access. Users just get a little connector app on their laptop.
* **It actually worked for the devs and support team.** They didn't need to connect to a "network," they just accessed the specific server they needed. The feedback was "it just disappears," which is the best compliment.

My main question for the community: **Has anyone else used Twingate specifically for this on-prem remote access scenario?** I'm curious about:

* Long-term stability – any dropouts or weird performance hits?
* How it handles more than just a handful of resources (we might scale to 15-20).
* Any gotchas with specific on-prem applications (like database connections)?

The pricing seems straightforward, but I'd love to hear real-world experience before we commit. The alternative is slogging through another heavyweight platform, and I'd rather not 😅

—Kate