Ask how two people are connected, what's happening around a target, or who on your team can reach them, and get the evidence behind it: real connections, shared history, recent signals. The same relationship data, whether a person, an agent, or your own product is asking.
One key. The same answers whether you use it through an AI assistant or your own software.
{
"data": [
{
"type": "path",
"id": "pth_8f2a",
"attributes": {
"degree": 1,
"layer": "private",
"hops": [{
"from": "sarah.chen@yourteam.com",
"to": "kevin.marsh@targetco.com",
"confidence": 0.92,
"evidence": [
{ "kind": "linkedinConnection", "verified": true },
{ "kind": "emailThread", "lastActivity": "2026-05" },
{ "kind": "meeting", "lastActivity": "2026-04" }
]
}]
}
},
{
"type": "path",
"id": "pth_3c7b",
"attributes": {
"degree": 2,
"layer": "public",
"hops": [
{
"from": "you",
"to": "david.okafor",
"confidence": 0.95,
"evidence": [{ "kind": "linkedinConnection", "verified": true }]
},
{
"from": "david.okafor",
"to": "kevin.marsh@targetco.com",
"confidence": 0.41,
"evidence": [{ "kind": "workOverlap", "company": "ezCater" }]
}
]
}
}
],
"meta": { "target": "kevin.marsh@targetco.com" }
}This is what the base account returns, every time: the strongest path first, the evidence behind it, and an honest read on how far to trust it.
Via ranks by how strong the relationship actually is, so the top result is the person whose connection is worth acting on, not just anyone who happens to be linked to the target.
Each path shows why it is real: a verified connection, a shared employer or school, a recent email or meeting. Not a mystery score.
Strong paths look strong and weak ones look weak. You get a clear read on every path, so you know which ones to actually use.
Via's relationship data is the base, and you get it by default with no setup. Connect your own network and it sits on top, prioritized, so your real connections rank highest.
Via's relationship data drops into what you already run and makes it warmer: the cold email that finds a real path first, the CRM record that shows the warmest way in, the account view your product ships with the receipts.
Your agent looks for a real path first. When one exists, it reaches out warm and cites the evidence: the shared employer, the recent thread, the person best placed to vouch. When one does not, it says so and moves on.
Score a target list, enrich a CRM record, and route the intro from the workflows your team already lives in. The evidence writes back to the record, so reps trust the score instead of ignoring it.
Show relationship strength and the warmest path on each account inside your own product, with the receipts that make the number worth trusting rather than ignoring.
Via returns the evidence behind a path, with confidence you can rely on. Here are the main things you can ask for.
Credits are only spent when Via returns a match. A lookup that finds nothing is free.
Full reference and quickstart at api.connectvia.ai.
Point Claude or your own AI assistant at Via (over MCP), or connect it straight to your software (REST). Either way you get the same results, so you never build it twice.
The answer is not a wall of text. An AI assistant can pull out one piece of evidence and use it in whatever it does next.
The docs are machine-readable (OpenAPI), so an AI assistant can read what is available and start using Via on its own, with no engineer to set it up.
Where our data comes from, what never crosses between customers, and how Via handles compliance and your data. We would rather claim less than overstate it.
Via licenses data from several outside providers and combines it with public information. That is what reaches beyond your own network: work and school connections we can piece together, never anyone's private connections.
No customer sees another customer's private graph. Customer-pooled data never enriches Via's public layer. Searches beyond your network return public data only: work and education overlaps and public signals.
SOC 2 is in progress with DSALTA. See the Trust Center for current status and documents.
To request removal of your data, see our privacy policy or write to privacy@connectvia.ai.
API access comes with your Via plan, and each plan includes monthly credits. A credit buys a warm path with the evidence behind it, not a row of data.
API access, bundled with the Via app you already use.
For teams building on the API or piloting an embed.
Give an agent a target, and get back the warm paths with the evidence behind each one.
Full docs live at api.connectvia.ai. Request access to get a key.