From Risk to Reason: Designing a Framework That Gave Automation Its Future

A Framework That Didn't Just Add a Feature—It Added a Future.

We launched a new AI integration framework. It was about giving our entire platform a new nervous system—one that could reason.

We established a governed, scalable foundation for AI that unlocked its power for the enterprise. It solved the security problem, democratized AI for everyone, and became a core driver of the business.

The Impact

  • Fueled Growth: Drove over 65% of new and upsell bookings in the year after launch.

  • Earned Trust: Reached 1 million+ intelligent automation runs within six months.

  • Shifted Time: Turned processes that took hours into minutes.

  • Delivered Value: Faster: A 3x acceleration in time-to-value for customers building with AI.

Project & Role

  • Timeline: 6 Months (July - Dec 2023)



  • My Role: Lead Designer. I was responsible for the strategic vision and end-to-end design execution. This meant defining the AI integration strategy, leading ideation, driving cross-functional collaboration (2 PMs, Research, Eng, Sales), managing and mentoring 2 other designers, and securing leadership buy-in for the new direction.

The Challenge:
When Rules Aren't Enough

The future of automation demanded more than rules. It needed to reason.

Our platform was a leader in telling software what to do. But integrating Generative AI was like fitting a jet engine onto a propeller plane. The core challenge wasn't just technical. It was a crisis of trust.

This isn’t just about a new capability. It’s about navigating a fundamental shift without breaking enterprise trust.

  • The Security Nightmare: For businesses, AI felt like a black hole for sensitive data. They had no control, no visibility. It was a non-starter.

  • The Complexity Barrier: For pro-developers, AI was a craft of bespoke, brittle integrations. It wasn't scalable.

  • The Wall for Everyone Else: For our Citizen Developers, AI might as well have been magic. Powerful, but completely out of reach.

The question wasn't if we should use AI. It was:

How do we build a bridge from the certainty of rules to the power of reason, and make it safe enough for everyone to cross?

My Process:
Seeing Past the Obvious Answer

It started with a simple ask. It ended with a new strategy.

Leadership asked for "AI templates." A simple, tactical request.
But the thing we often miss is that the easy answer rarely solves the real problem. My research pointed to a deeper truth. We didn't just listen; we immersed ourselves:

  • Internal Deep Dives: Talked with our expert developers and solution architects.

  • External Listening: Scoured developer forums and analyzed competitor approaches.

  • Cross-Functional Workshops: Mapped customer pains and platform dependencies with Sales, Services, and PMs.



The insight was clear: templates were a patch.

The real pain was systemic. Admins couldn't govern. Developers couldn't scale their work. A template would only add to the chaos.

The AI landscape was evolving daily. Waiting for direction wasn't an option. I took the initiative to become the team's subject-matter expert, curating a channel with key industry shifts, competitive analyses, and emerging AI capabilities. This proactive research didn't just inform our project; it directly influenced the wider product roadmap and helped us see what was next.

This wasn't a UI problem. It was a systems problem.



Based on the research, I developed and championed a new vision built on a simple, three-part mental model that served each persona's core need:

  • Govern (For the Admin)

  • Create (For the Pro-Developer)

  • Consume (For the Citizen Developer)



This reframed the entire project. I didn't win them over with data. I told them a story of two developers: one who failed with a complex template, and another who succeeded with a simple, pre-built "AI Skill." That story made the vision tangible. It made the value real.

The Solution:
A Framework for Clarity and Trust

Our final design brought this three-part vision to life in a dedicated "AI Studio."

  • For the Admin: Governance Brings Control. A central command center to approve and manage every AI model. This turned the black hole of risk into a glass box of control, unblocking our biggest customers.

  • For the Pro-Developer: Creation Enables Scale. A dedicated studio to build reusable "AI Skills." This turned a developer's niche expertise into a scalable asset for everyone else.

  • For the Citizen Developer: Consumption Creates Access. A simple, drag-and-drop experience. We made AI feel obvious, lowering the barrier to entry from "expert" to "everyone."

A dedicated space to build reusable "AI Skills." This turned a developer's niche expertise into a scalable asset for everyone else.

A simple, drag-and-drop experience. We made AI feel obvious, lowering the barrier to entry from "expert" to "everyone."

A central command center to approve and manage every AI model and AI skills.

From Vision to Reality:
Shipping the Solution

We worked in agile sprints, using regular design reviews to ensure alignment. We didn't wait until the end to test. We used interactive prototypes with internal and external developers to find friction early.

One key insight came from this testing: users were confusing our new "AI Skills" with standard automation actions. It wasn't clear enough. This led directly to refining the UI—improving iconography and labeling within the builder—to create an obvious distinction. That small change made a big difference.

Project launch and media coverage

Reflection:
Lessons in Crafting Strategy

This project was a lesson in strategic leadership. The numbers tell one story; the shift in possibility tells the rest.

  • Challenge the Premise: The first ask is rarely the right one. Our role is to find the question behind the question. True leadership isn't just executing; it's ensuring the vision is sound.

  • Abstraction is Empathy: By creating distinct layers—Govern, Create, Consume—we met each user where they were. We absorbed the system's complexity so they didn't have to. That’s the real work of design.

  • Ship to Learn, Craft to Last: We launched an MVP, but the vision was always bigger. We built the foundation first. This framework wasn't just an update; it was the groundwork for the company's agentic AI future.