Designing the future of intelligent workflow automation
Transforming rule creation from a technical bottleneck into an intuitive visual operating system that is built for scale, owned by everyone, and far harder to outgrow.I sprung rule creation out of the code dungeon and turned it into a visual operating system anyone can play with. Built to scale, owned by everyone, and a lot more fun to use.
Enterprise rule engines are broken by design
The tools that run business-critical automation were quietly failing the people who depend on them most. Logic lived in code, so operations teams couldn't touch it. Dependencies went undocumented, so workflows broke in production for reasons nobody could see. The pain wasn't cosmetic: it was structural, and it was expensive.
Baseline figures from the project's discovery phase.
An operating system for business decisions
Not a workflow tool. Not another rule engine. The bet was to treat visual logic as a first-class product: readable, scalable, and owned across the organisation instead of locked inside engineering. The market kept offering the same false trade: tools simple enough for business users were too shallow for the enterprise, and tools powerful enough for the enterprise were unusable for the people who actually run operations. This closed that gap.
Making complex logic feel simple
A visual node canvas
Logic flows left to right like thinking does. Drag-and-drop blocks map to real business concepts, and connective lines make every dependency explicit and auditable by anyone.
Progressive disclosure
Beginners land in a guided, templated space. Power features (branching, nested conditions) appear only as a user's actions signal readiness. The interface teaches as it expands.
AI as infrastructure
Embedded reasoning, not a chatbot: it recommends the next node, flags contradictory rules before production, and translates any workflow into plain-English anyone can read.
A system, not screens
Every component was decided at the system level: governed primitives, shared states, tokens. Adding a new surface takes minutes, not weeks, and stays consistent by construction.
Design that moved the metrics that matter
Workflow error rates dropped measurably once visual validation rails and conflict detection shipped, translating straight into fewer production incidents.
The interface was the intervention: it didn't just build rules faster, it handed business teams ownership of their own logic.