The Glass vs. The Water
Think of your SaaS stack like a collection of expensive glassware.
You've spent years buying the finest crystal: your CRM, your project management tool, your communication platforms. But the value isn't in the glass. The value is in the water, the data inside.
For years, SaaS companies have thrived by making their "glass" opaque. They want you to believe you need their specific UI, their reporting logic, their "AI-powered" features just to access your own data. They want you to believe that "Pipedrive is how we sell" or "Linear is how we build."
The result is what I call the SaaS Tax: you pay more every year for features you don't use while your data remains trapped in silos. Worse, you hire people to act as human bridges, coordinators whose job is to manually move information from the CRM glass to the project management glass to the communication glass.
We decided to stop paying that tax.
What We Actually Built
At FM, we still use Pipedrive for CRM and Linear for project management. We're not religious about tools. But we stopped treating them as the center of our operations. Instead, we built an agentic layer that sits above everything.
Here's what that looks like in practice:
Proposal Generation pulls meeting notes and Pipedrive deal context to create branded PowerPoint proposals. No one manually assembles slides anymore.
Meeting Synthesis takes raw meeting notes, structures them into documentation, and commits them to our knowledge base automatically. The insight doesn't die in someone's notebook.
Weekly Synthesis generates client updates by pulling from meetings, Slack conversations, and Linear tickets. It creates internal docs and drafts client emails. What used to take our team hours of Monday morning work now happens automatically.
SOW Builder generates Statements of Work with consistent language, type-specific templates, and dual storage to Google Docs and GitHub. Same quality every time, fraction of the effort.
Company Research takes a prospect name, runs web research, pulls CRM context, and generates a tailored sales call script. Our team walks into calls prepared without the prep work.
Content Brainstorm helps us generate newsletter content with FM's voice, incorporating web research and exporting directly to our editor.
Each agent crosses system boundaries that used to require human coordination. The Proposal Generation agent doesn't care that deal data lives in Pipedrive and meeting notes live somewhere else. It just assembles what's needed.
The Shift in How We Think About Software
This changed how we evaluate tools. We now buy based on one criterion: the API. If a tool doesn't have a solid API, we won't use it. The UI is almost irrelevant because our team increasingly interacts with data through our own interface, not the vendor's.
It also changed how we think about "AI features" inside SaaS products. Most vendors are currently bolting on AI and charging $30/user/month for features that summarize your notes or suggest next steps. That intelligence is locked inside their glass. We'd rather build intelligence that sees across all our systems at once. That's where the leverage actually lives.
When This Makes Sense
I want to be honest about the threshold here. Building your own operating layer isn't always worth it.
It makes sense when: You have at least 3-4 core systems that need to talk to each other. You're paying for coordination labor (even if you don't call it that). Your workflows have enough repetition that automation compounds. And you have access to someone who can build and maintain custom integrations.
It doesn't make sense when: You're a team of five people who can just talk to each other. Your tools already integrate well enough through native connections. The coordination cost is low. Or the build cost exceeds several years of the "tax" you're currently paying.
For FM, the math was clear. We were spending real hours every week on work that was essentially moving water between glasses. The agents we built paid for themselves within months.
The Bottom Line
The "System of Record" mentality made sense when software was expensive and scarce. You picked your platform, committed to it, and built your processes around its constraints.
That era is ending. When building custom integrations costs a fraction of what it used to, the smart move is to treat your SaaS tools as replaceable databases with APIs and build your own logic on top.
You don't need to replace your tools. You need to stop letting them dictate how you work.
At FM, we help companies build operational intelligence through AI and systems thinking. If you're ready to move your AI agents from prototype to production-grade operations, let's talk.
