Cofia Builds Your Automations Before You Know You Need Them

The Macro: Automation Is Great, but Nobody Has Time to Set It Up

The dirty truth about workflow automation is that most people who would benefit from it never set it up. Zapier has millions of users but the vast majority of knowledge workers have never created a single Zap. Make (formerly Integromat) is powerful but requires you to think in flowcharts. n8n is flexible but demands technical confidence. The bottleneck is not the capability of the tools. It is the effort required to get from “I do this repetitive thing every day” to “a machine does this repetitive thing for me.”

I have watched countless colleagues manually copy data between apps, manually send the same follow-up emails, manually update spreadsheets from Slack messages, all while knowing that automation exists. They just never get around to building it. The setup cost is too high. You need to identify the pattern, open the automation tool, connect the services, configure the triggers, test the flow, and maintain it when something breaks.

The automation market is massive. The workflow automation segment alone is projected to exceed $30 billion in the next few years. But the penetration rate among potential users remains low because the tools are built for people who already think in systems. They are built for the kind of person who draws process diagrams for fun.

Cofia is trying to flip this entirely. Instead of you building automations, Cofia learns how you work and proactively creates tailor-made automations that you can deploy without ever writing a prompt.

The Micro: It Watches, It Learns, It Suggests

Cofia’s approach is observation-first. The product watches your work patterns, identifies repetitive tasks, and generates custom automations tailored to your specific workflows. You do not need to describe what you want automated. You do not need to scroll through templates. You do not need to know what a trigger or an action is. Cofia figures it out.

The “without you ever writing a prompt” angle is the key differentiator. Every other AI-powered automation tool still requires you to articulate what you want. Cofia claims to skip that step entirely. It watches, it proposes, you approve or reject. The automation writes itself.

This is a fundamentally different user interaction model than what Zapier, Make, or any existing automation platform offers. It shifts the cognitive load from “what should I automate and how” to “does this suggested automation look right.” That is a much simpler decision for most people.

Paola Martinez and Moses Wayne are the cofounders. Martinez led retention strategy at Brilliant.org and has a Stanford background. Wayne directed monetization at Duolingo and studied at Duke. Both come from consumer product backgrounds where understanding user behavior at scale is the core competency. That is relevant because Cofia’s entire value proposition depends on understanding user patterns accurately enough to generate useful automations. The company came through Y Combinator’s W26 batch.

The risk is obvious: observation-based automation needs access to your workflow data, and the privacy implications are non-trivial. How much does Cofia see? Where does that data go? These are questions enterprise buyers will ask immediately.

The competitive moat, if it works, is strong. Zapier and Make would need to fundamentally rethink their product model to match this. They are template-first tools trying to become smarter. Cofia is intelligence-first, trying to generate templates. That is a hard pivot for an incumbent.

The Verdict

Cofia is making one of the boldest bets in productivity tooling right now. If observation-based automation works, if the system can accurately identify patterns and generate useful workflows without user input, it unlocks the 90% of knowledge workers who never build automations themselves.

At 30 days: how good are the suggestions? If Cofia suggests ten automations and only one is useful, users will stop paying attention. Precision matters more than recall here.

At 60 days: how does it handle false patterns? Everyone has one-off tasks that look repetitive but are not. The system needs to distinguish between “you do this every Tuesday” and “you happened to do this two Tuesdays in a row.”

At 90 days: the privacy story. Enterprise adoption depends on clear, auditable answers to what data Cofia observes, how it stores that data, and who can access it.

The Duolingo and Brilliant backgrounds are strong signals for a product that depends on understanding human behavior at scale. I think Cofia is onto something real, but the execution bar is extremely high.