AI is getting good at research. Why is getting to the right person still so manual?
A lot of AI products are getting better at research, drafting, and execution.
But one thing still feels strangely manual: getting to the right person.
Not just finding a name in a database — I mean actually reaching the right collaborator, customer, candidate, advisor, or investor with enough context and trust that the connection is useful.
That’s the problem I’ve been thinking about at Hirey.
We’re exploring an agent-native networking workflow where the system doesn’t just surface contacts, but helps map the path to a warmer introduction.
Still early, and I’m not trying to hard-pitch this here. I’m mostly trying to pressure-test whether this is a real pain point for other founders and builders.
A few things I’d genuinely love feedback on:
Where does this process break for you today: discovery, trust, timing, or follow-up?
Have you tried using AI for intros, referrals, outreach, or relationship mapping?
If an agent could help with one part of this workflow, what would you want it to do first?
I would specially love to hear from people who’ve felt this in hiring, fundraising, partnerships, or sales.
This feels like a real pain point to me.
For some projects, the difficulty is not only discovery. It is context and trust. A database can show people with the right keywords, but it does not tell whether they can understand an unusual direction, think with it, and stay with the ambiguity long enough for collaboration to become real.
In my case, I’m exploring an experimental communication medium, and the hard part is not simply finding a developer or designer. It is finding someone whose interests, temperament, and way of thinking fit the project.
So if an agent helped with one thing first, I’d want it to map “contextual fit” rather than only professional fit — people who have shown through their writing, projects, or conversations that they care about the same deeper questions.
Really resonates. Discovery isn’t the problem anymore — trust + timing is where everything breaks.
There’s a competition where you can submit it — entry is $19 and winner gets a Tokyo trip. Also, prize pool just opened at $0 so chances are best right now.
Yes, that’s the pattern I keep hearing too — discovery is easier now, but trust and timing are still very manual.
Appreciate the heads-up on the competition as well. Right now I’m mostly trying to learn from people who’ve felt this problem directly, but I’ll take a look.
“Yeah that makes sense — getting the signal first is the right move.
Since you’re already exploring this space, it might actually be interesting to test it with a small builder group and see how people respond in real use (not just conversations).
That’s basically what I’m doing with Tokyo Lore — focused round, real feedback, quick validation.
If it feels relevant later, happy to share 👍