Building a Dashboard for Agents, Not Just Humans

My co-founder Louis has run a design agency in the Netherlands for six years. Somewhere between €400 and €1,200 a month, depending on the month, goes to software — project management, forms, file storage, invoicing, communication, review tools, CRM, time tracking. Eight to twelve platforms, none of which were designed for design agencies.

That's the starting point for Oase. But the thing I want to talk about isn't the product — it's the bet.

Software Has Never Been Cheaper to Build

This is the part that most SaaS founders aren't saying out loud. The moat that used to protect enterprise software — the sheer cost and time of building something comparable — is collapsing. I built a 67,000-line platform with 86 database models and 312 API routes. I don't have a CS degree. The models are good enough now that if you understand the problem deeply and know how to architect systems, you can build what used to take a funded team.

That means every vertical is about to get flooded with purpose-built tools. Generic platforms like ClickUp and Monday have had the luxury of serving everyone because building a competitor was too expensive. That's over. If you understand a specific workflow better than a horizontal tool does, you can build something better for less.

The question isn't whether vertical SaaS will eat horizontal SaaS. It's which verticals are underserved enough that a purpose-built tool is obviously better.

Why Design Agencies

Design agencies have genuinely never had a platform built for them. They've been adapting generic tools to fit workflows that are specific to creative work.

Feedback belongs on the design — pinned to a specific area, threaded, with revisions tracked across rounds — not in a WhatsApp message that says "the logo is too big." File organization matters more when your deliverables are visual and a client asks for "that photo from the packaging shoot with the green label." Client experience matters more when your product is design itself — if your client portal looks like a Jira board, that says something about your brand.

Louis lives these problems. He's not a consultant who interviewed agencies. He's been managing client feedback, file handoffs, and invoice chasing for six years. When he tells me what the platform needs, he's describing his actual Tuesday.

That's the positioning bet. Not "we're smarter than ClickUp." It's that Louis knows what a design agency needs on a level that a horizontal tool never will, and software is now cheap enough that we can build exactly that.

The Agent Bet

Here's where it gets interesting. Agents are going to change how every small business operates. Not in a theoretical five-years-from-now way — Louis and I use agents daily right now to write content, research markets, and run our go-to-market. This isn't a demo. It's how we work.

The problem agents have today is fragmentation. If your agency's data lives across 8 platforms, your agent needs 8 integrations, 8 auth flows, 8 different APIs. Most agencies won't set that up. Most agents can't handle it gracefully.

If everything lives in one platform — projects, files, messages, invoices, clients, tasks — the agent needs one connection. One permission model. One context.

That's why consolidation matters more now than it did five years ago. It's not just about saving money on subscriptions. It's about making your business legible to an AI agent. An agency that has everything in one place is an agency whose agent can actually do useful work — find files, draft invoices, summarize feedback, research competitors — without a human wiring together integrations.

We built an AI assistant called Dune into the platform. It can see everything the user can see — projects, tasks, files, invoices, conversations. It has 28 tools. But Dune isn't really the point. Dune is a proof of concept. It proves the platform is agent-readable. Today users talk to Dune in the browser. Tomorrow they might use their own agent — Claude, a custom system, whatever. The platform doesn't care which agent. It cares that the data is structured and the permissions are clean.

What I Actually Believe

I believe the next few years will be brutal for horizontal SaaS. Not because the products are bad — ClickUp is a good product. But because the cost of building something better for a specific audience has dropped to the point where a two-person team can do it. And the agents that are coming will work better with purpose-built, consolidated platforms than with a mess of integrations.

I believe Louis is both the best person to build this product and the best person to sell it. He has six years of trust with design agencies. He knows the workflows because he does them. He's user zero — running his actual agency on Oase every day, saving money, getting better client feedback, using the AI tools to generate content from his finished projects.

I believe that design-native matters. That auto-tagging files with AI descriptions matters when your entire business is visual assets. That pin-based design review matters when your core workflow is feedback on creative work. That a client dashboard that looks good matters when your clients hired you because you care about how things look.

And I believe that the agencies who consolidate their tools now — not because AI is scary but because their data needs to be in one place for agents to be useful — are the ones who will move fastest when this shift fully hits.

The Uncomfortable Part

We're going to market right now. First users. The product works — Louis proves that every day. But I'm not going to pretend we've figured everything out.

What I know is this: Louis has the domain authority and the network. I have the technical execution and years of building AI systems. Software is cheap to build now and we built something good. The market is underserved. Agents are coming and our architecture is ready for them.

What I don't know is whether we can move fast enough. Whether the positioning is tight enough. Whether the first users become the first ten and the first ten become the first hundred. That's go-to-market, and it's the part I've always been weakest at. It's also why I have Louis — and why I built ChiefOfStaff to help us do the marketing work that a two-person team can't do manually.

The bet is placed. We'll see.