Right now, when your AI needs to call Shopify, Notion, Slack, GitHub, and Gmail — you either give it your API keys (risky) or set up OAuth for each one (exhausting). We've been through this. We know the ceiling it hits.
We're looking for other developers who have hit the same wall.
Does your AI agent need to call more than one external platform on behalf of a user?
Not just one — we're asking about the moment you needed a second, third, or fifth integration.
Question 3 / 10
Today, when your AI needs access, the standard approach is OAuth 2.0 Device Flow + passing the user's API Key / token to the agent. We've done this too. The question is — what happens when your agent needs to interact with 3, 5, or 10 different platforms?
Be honest — have you hit this ceiling yet?
Question 4 / 10
When your AI needs to talk to different services (Shopify API, Slack API, Notion API, GitHub API, Gmail API), how do you handle the fragmentation?
Select all that apply — from what we've seen, most teams end up with a mix.
Select all that apply, then click Next.
Question 5 / 10
How much do you agree? "The problem isn't 'how do I authenticate my AI' — it's 'how do I give my AI a unified identity that works across every platform it needs to touch.'"
Question 6 / 10
Here's a question we've been sitting with:
"Every SaaS platform has user accounts — for humans. But when an AI agent calls your API, it borrows a human's account. The AI itself has no identity. If AI agents are becoming your most active 'users,' shouldn't they have their own identity?"
Has this thought crossed your mind?
Question 7 / 10
If AI agents were to have their own identity — separate from the human's — how should that work ideally?
Select the option closest to your ideal.
Question 8 / 10
What worries you most about AI agents getting their own identity?
Select the biggest one.
Question 9 / 10
Last one — we'd love your honest read:
"AI agents need a first-class identity — not borrowing a human's API key, but a proper digital identity that platforms can recognize, scope, and audit. Today that doesn't exist, and it's the bottleneck holding back autonomous AI agents."
Question 10 / 10
If you could give your AI agent its own identity tomorrow — one login, any platform — what's the first thing you'd let it do for you?
No right answer. We just want to know what you'd actually do.
Want to stay in touch? (optional)
We'll only reach out if we have follow-up questions or share the results.
✨
Thank you
You've helped us understand whether AI agents need their own identity — or whether borrowing a human's credentials is good enough.