Understanding inferred connections

Most of what Affinity knows about your network comes from explicit relationships — people you’ve actually emailed, met, or scheduled — real activity. But on a new deal, it often dead-ends: you’ve never contacted anyone at the target, so any path stays invisible. Inferred connections fill that gap. An inferred connection is:

• A likely relationship Affinity surfaces from shared history,
• even when no one at your firm has interacted with that person directly, and
• a lead worth checking — not a confirmed relationship. (Think of two people who attended the same college in the same years: they probably crossed paths, but you can’t be sure.)

​

The two types in this release

• Overlapping work history — “your teammates likely know them.”
  • A contact worked at the same company, over an overlapping stretch of time, as one or more of your teammates.
• Investor ↔ portfolio-company executive — “known through your firm’s network.”
  • Leverages the first second-degree, external-to-external connection Affinity builds.
  • A contact your firm already knows was at an investment firm when it funded a target company, so they very likely know that company’s executives and founders.

​

Where you’ll see them

For this release, inferred connections appear in one place:

• Open any Company Profile → Connections tab.
• Below your direct connections, look for “Inferred Connections to {Company}”, with a sub-header like “3 people likely within your firm’s network.”
• Each row is a connection card — one card per person your firm might be able to reach at that company.

​

Reading a collapsed card

By default each card is collapsed and shows:

• The person — name, title, and company (e.g. “John Smith · CTO @ Potluck”). A LinkedIn icon appears when we have a match.
• One or two relationship rows explaining why they surfaced:
  • Work-history row — e.g. “5 teammates have overlapping work history.”
  • Funding row — e.g. “Carol Channing was a Partner at Greenleaf Investing when it funded Potluck under CTO John Smith.”
• When both rows appear, work history sits on top — a teammate who actually worked alongside someone is usually a warmer path than a funding-era overlap.

​

Expanding a card for the full timeline

Click a funding / investor-executive card to expand it in place into a timeline view:

• A horizontal bar for each person across the relevant years.
• A shaded band marking the window they shared.
• A dashed marker on the funding round (e.g. “SERIES B”) with a plain-language summary — “Carol Channing was an investor at Greenleaf Investing 8 months ago when Potluck raised its Series B on March 25, 2025.”
• A ⋯ menu on each person’s row to view their profile, open LinkedIn, or copy their email.
• Only one card is open at a time — opening another closes the first.

​

Find a warm intro with AI Chat or MCP

You can also reach inferred connections through AI Chat and the Affinity MCP — just ask the way you’d ask a well-connected colleague:

• “Who can introduce me to someone at Crusoe Energy?”
• “Do we know anyone at [investment firm]?”
• “How can I get introduced to the CEO of [company]?”

How the results are presented:

• Ranked warm paths come first — your real, interaction-backed relationships.
• Inferred connections appear separately, in a clearly labeled “possible paths” section — never blended into warm paths, and never called “warm.”
• Each one shows the basis of the inference in plain language — “overlapped at Jobbers 2018–2021” or “both connected to Potluck via Greenleaf Investing” — so you can judge it. There’s no invented strength score; the basis is the value.
• Empty result? It means none surfaced for that company — not an error, and not that you have zero. Coverage varies by company and by your firm’s access.

​

How Affinity infers an investor ↔ executive connection

You’re being handed a prediction, so it helps to see the chain behind it:

1. A funding event links an investor firm to a company it backed.
2. Affinity checks whether a person your firm knows worked at that investor firm during the investment period — at the time of the deal, not just sometime in their career.
3. It filters for relevant roles — partners and investors on the investor side; C-suite, founders, and board members on the company side — so the bridge is senior enough to plausibly know the target’s leadership.

Because the match is anchored to when someone was somewhere, it’s a historical overlap — not a loose “same industry” guess.

• Affinity scores each inferred connection for confidence (longer, more recent overlaps and smaller companies score higher) to decide what surfaces and in what order.
• You won’t see a confidence number — it works quietly in the background, the same way relationship strength does for your explicit relationships.

​

What’s not available yet

• Company profiles only — not yet on individual people’s profiles (including bridge-contact profiles). Fast follow.
• No click-through to an unknown target person’s own profile page yet — gated on a larger data effort.
• Not in global search yet — searching a target’s name won’t surface them as inferred (same data gating).
• The Connections-tab number badge counts direct connections only — it doesn’t include inferred ones yet, so the badge may look low while inferred matches appear in the section below it.
• Work-history cards can’t be expanded into a timeline at launch (see the note above).
• Not in this release: mobile, the connection side panel, and the Chrome “Path Finder” extension.

​

FAQ