← back to all posts
July 29, 2025

Building a Discovery Protocol

youtube

Can discovery happen without exposure? Hear from Seref Yarar, Index Network Co-Founder in conversation with Chad Fowler, BlueYard Capital GP & CTO.

Search no more

We've spent two decades training ourselves to think like search engines. Pick the right keywords. Filter by category. Scroll through results. Click, evaluate, repeat.

But what if the fundamental metaphor is wrong? What if connection isn't something you search for but something you declare?

This is our philosophical premise. Instead of users actively hunting for connections - like co-founders, collaborators, communities - they articulate what they want, and AI agents work in the background to surface mutual fits.

Very soon, instead of going to search engines where you pick what you search for, there will be these agents working in the background connecting people who share mutual intent.

The three flavors of intent

Not all intents are created equal. There are three types:

Information-seeking intents are what we're most familiar with: "What's New York's population?" These are the bread and butter of search engines, discrete questions with discrete answers.

Financial intents move into declarative territory: "I want to buy Bitcoin at $115,000." You're not searching for sellers; you're stating terms and letting the market match you.

Social intents - Index's focus - operate in a murkier space: "I'm looking for a technical co-founder in climate tech" or "I want to launch a community around AI ethics." These are fuzzy, multi-dimensional, and historically resistant to automation.

The shift from search to intent isn't just semantic. It's about moving cognitive load from humans to machines. You shouldn't have to translate your actual desire ("I need help validating this business model") into platform-legible behaviors (posting on LinkedIn, joining Slack groups, cold emailing 50 people).

Retiring from self-advertising

Current platforms solve discovery through broadcasting. Optimize your profile, engage with the algorithm, hope the right people see you.

Advertising is when you stake on yourself. But this creates an arms race of visibility where everyone's shouting and no one's listening.

The practical toll is brutal: post a job on LinkedIn, pay $200-1000, receive 1000 applications. 950 are auto-generated spam. You're left with 50 candidates to manually evaluate using the platform's limited filters (location, years of experience, skills) - categories that rarely capture what actually matters.

Dating apps, professional networks, community platforms have weirdly all converged on this same exhausting pattern.

Mutual intent as the coordination mechanism

Index introduces a different organizing principle: mutual intent as the basis for connection.

An agent might say: "You should invest in this company." Or: "Five of you should launch a community together." Or: "You and this person should work on this problem."

These aren't random matches. They're based on complementary declared intents, validated through double opt-in. Both parties have to say yes. Agents can vouch for connections, but neither side is forced into visibility they didn't choose.

This distinction matters. You're not broadcasting yourself to everyone hoping someone notices. You're declaring what you actually want, and the system routes you to people who want complementary things - when both of you are ready.

Privacy as infrastructure

Here's where it gets technically interesting: how do you match people on nuanced, multi-dimensional intents without exposing all their data?

Index uses contextual privacy primitives to ensure appropriate information flows. Your detailed intents—the full context of what you're building, what you need, what matters to you—stay private. But agents can still compute compatibility and suggest connections.

As Chad says it: "Privacy isn't just about hiding information, but it's also about being able to use your most valuable or personal information to make things better for you and for others."

Privacy becomes infrastructure for richer matching, not a constraint that limits it.

From searching to declaring

What becomes possible when discovery stops being a search problem?

You could declare: "I'm raising a seed round for climate infrastructure, need technical advisors, and want to connect with other founders thinking about category creation."

Agents could work across platforms - LinkedIn, Twitter, GitHub, community Slacks - finding people with complementary intents. Instead of you manually checking 10 places, agents surface: three potential investors actively looking for climate deals, two advisors who just posted about wanting to support early-stage founders, and a founder in an adjacent space who's been thinking about the same category questions.

All of this happens in the background. You get introduced when there's actual mutual fit, not when you happen to post at the right time in the right place.