How Costanoa Ventures turned specialized networks into a scalable warm-intro engine
“Swarm is allowing us to rethink relationship management. We’re moving beyond LinkedIn mutual connections to a much broader understanding of who actually knows whom, and how to turn those relationships into warm introductions.”
Jim Wilson, Operating Partner, Costanoa Ventures
The challenge
For a venture firm, relationships are one of its most valuable assets. But the network itself is only useful if the team can quickly identify the right person to ask for the right introduction.
Costanoa Ventures supports more than 100 portfolio companies across business development, talent, and company building. The firm had access to a large and highly valuable network, including specialized groups of advisors and operators, but finding the strongest path into a target company often meant manually searching LinkedIn and piecing together mutual connections.
That approach did not scale, especially when the most valuable path might sit outside the immediate Costanoa team.
Building purpose-specific networks with Swarm
Rather than treating its network as one large undifferentiated graph, Costanoa began creating specific Swarm teams around strategically important communities.
For example, the firm has built a dedicated cybersecurity Swarm made up of advisors and operators who are particularly well connected across the industry.
When a portfolio company needs an introduction into cybersecurity, Jim Wilson does not start by searching the entire Costanoa network. He can go directly to that specialized Swarm and identify the people most likely to provide a credible path.
This gives Costanoa a way to organize relationship intelligence around the communities that matter most to its portfolio.
Turning Swarm data into a Clay workflow
Costanoa then built a workflow in Clay that takes the process several steps further.
A portfolio company provides a list of target companies or people. The Clay workflow pulls relationship data from the relevant Swarm, analyzes the available paths, and helps surface the strongest potential connector.
From there, the workflow can continue all the way through to drafting the outreach email asking for the introduction.
The process looks roughly like this:
- Start with a target account or target person.
- Query the relevant specialized Swarm.
- Pull the available relationship paths into Clay.
- Sort and prioritize the strongest connections.
- Identify the best person to ask for the introduction.
- Generate a draft email with the relevant relationship context.
- Have Jim review the recommendation before sending.
The result is a highly automated workflow without removing human judgment from the final introduction.
For Jim, who helps Costanoa’s portfolio companies reach prospective customers, this has become a repeatable way to activate relationships at scale. Today, he facilitates approximately 100 warm introductions per quarter.
“How many people have spent the last few years looking on LinkedIn and just looking at mutual connections? You just don't know. No one's ever really been effectively able to crack second- and third-degree connections and answer: how do I actually reach someone to get a warm intro?”
Jim Wilson
From LinkedIn search to relationship infrastructure
The important shift is that Costanoa is no longer relying on one person’s memory or a simple LinkedIn mutual connection.
Swarm lets the firm build distinct relationship networks, bring those networks into Clay, and turn them into an operational workflow.
That means a portfolio company can move from:
“Does anyone know someone at this company?”
to:
“Who is the strongest person in our cybersecurity network to make this introduction, and what should we ask them?”
Expanding into Claude with MCP
Costanoa is now taking the same relationship data into AI workflows. The team is adopting The Swarm MCP integration for Claude, allowing team members to query and work with their relationship network directly inside Claude.
Instead of manually navigating a network map, they can begin using natural language to explore questions such as:
- Who can get us into this account?
- Which advisor has the strongest path to this person?
- What are our best routes into these 20 companies?
- Who should we ask for an introduction?
This gives Costanoa another interface for the same underlying relationship intelligence, particularly as the firm adopts more AI-native internal workflows.
Bringing warm intros into Attio
Costanoa also uses Attio as part of its CRM stack.
Through The Swarm’s Attio integration, the team can surface warm introduction paths directly alongside the companies and people it is already tracking.
That reduces the need to switch tools just to answer a relationship question. Instead, relationship context can appear where the team is already managing sourcing, portfolio support, and outreach.
One relationship layer, multiple workflows
What started as a way to better understand Costanoa’s network has evolved into a relationship-data layer used across several workflows:
- Specialized Swarms for high-value communities such as cybersecurity
- Clay workflows that identify the best connection and draft intro requests
- Portfolio GTM supporting approximately 100 warm introductions per quarter
- Claude MCP for natural-language access to relationship intelligence
- Attio for surfacing warm paths directly inside the CRM
- Deal sourcing and talent using the same underlying network data
“The Swarm unlocks high-impact intros for our team and founders.”
Katy Wiley, Director of Talent, Costanoa Ventures
The result
Costanoa has moved from manually searching individual social graphs to systematically activating purpose-built networks.
The firm can now organize valuable communities, identify the strongest path to a target, bring that intelligence into Clay, generate the introduction request, and increasingly access the same network through Claude and Attio.
Swarm is becoming less of a destination for Costanoa and more of a relationship intelligence layer across the firm’s existing stack.
