TL;DR
HubSpot’s Cody Holden joined Stefan to share his story from newsrooms to the frontier of API integration. The conversation explored how a foundation in writing became the backbone for his product work, why communication defines developer experience, and why MCP (Model Context Protocol) is becoming an important part of how SaaS products connect AI systems with APIs.
No matter how valuable something is… if I’m unable to communicate that to an audience, it really diminishes the value.
From journalism to product management
Cody’s story starts in Baton Rouge, Louisiana, where he studied at LSU and worked as a news editor covering mayoral campaigns, local businesses, and restaurant culture.
He said the newsroom taught him to communicate clearly under pressure; skills that carried into product management.
Being able to communicate with different audiences of varying technical expertise is what gives any product its real value.
That grounding in clarity and empathy would later become the lens through which he designs APIs and developer tools.
The first step into tech
After journalism, Cody joined MitraTech, a legal-matter-management company, as a technical writer.
At the time, documentation was shipped to customers as PDFs, which quickly became outdated. Cody helped move the company’s documentation online so customers could always access the latest version.
If I give someone a document, they’re not getting the latest updates on it… they’ll be out of date and out of compliance.
He realized technical writing was about more than words. It involved creating systems to keep information current and useful, which led him toward product management.
At HubSpot: connecting AI and APIs
Today, Cody leads MCP and Integrations Platform at HubSpot. MCP (Model Context Protocol) is HubSpot’s bridge between software and AI systems, allowing models to safely call APIs on behalf of users.
MCP is about making APIs understandable and usable by AI, safely.
He explained that every SaaS product is now wrestling with the same question: how to give AI access to data without compromising security or context. He described HubSpot’s approach as treating MCP as part of the overall developer experience, tightly integrated with API governance and tooling.
Cody compared his earlier focus on helping people understand APIs to now helping AI systems safely understand and use them through MCP.
Communication as the core of good engineering
Stefan and Cody both agreed that the hardest part of building great software isn’t code—it’s communication. Engineers must communicate with users, with teammates, and increasingly with AI systems that act on their behalf.
For Cody, the bridge between news writing and product management isn’t as wide as people think.
Whether you’re writing news or defining an interface, the goal is to make complexity digestible.
That belief drives his approach to developer experience at HubSpot: if you make something easy to understand, you make it powerful.
Looking ahead: the evolution of integration
Toward the end of the episode, Cody and Stefan discussed how MCP represents a shift in integration by moving from simple API calls to AI systems that can act across multiple products. Cody emphasized that HubSpot’s investment in MCP doesn’t replace existing integration patterns; it complements them.
REST APIs remain foundational, and new AI-focused protocols like MCP are being built alongside them.
The result is that MCP could make APIs easier for AI systems to find and use safely, without replacing existing standards.
The bigger picture
Cody’s path to managing MCP and integrations at HubSpot reflects a broader story about how communication shapes modern software. Whether it’s journalists making news accessible or developers making data interoperable, clarity is what connects people, systems, and ideas.
As AI begins to inhabit every layer of the stack, Cody emphasized that clear communication should come first when building software.
This episode was directed by Jacob Javor. Transcript lightly edited for clarity and flow.
