SYSTEMS THINKING
Duration
April 2026 - June 2026
My Role
Drove product strategy and end-to-end UX & workflow design to prototype using Claude Code.
OVERVIEW
Designing a centralized platform for governing and controlling AI usage across the enterprise
As AI systems become increasingly embedded in enterprise workflows, organizations need a structured way to govern how they are used, monitored, and controlled. GovernAI brings together policy definition, and automated enforcement into a unified experience.
PROBLEM
AI is becoming embedded in everyday work, but most organizations still lack a clear way to govern how it is used
AI usage often spreads faster than governance processes can keep up. Teams are left managing critical questions around visibility, permissions, approvals, and policy enforcement without a unified way to oversee and control AI activities across the organization.
Who is using AI, and for what purpose?
What info can be safely shared with AI?
How can governance be enforced?
INSIGHT
Governance ultimately comes down to policies, yet creating and managing those policies remains complex & fragmented
Organizations may understand the risks of AI usage, but they often lack intuitive ways to define guardrails, encode governance decisions, and test policies before they go live.
User Quotes from Interviews
GOAL
What if creating AI policies was as intuitive as building workflows?
THINKING IN SYSTEMS
Translating AI governance into a repeatable decision-making system
AI governance is not a linear process. Policies need to be created, tested, enforced, and continuously refined as new use cases emerge. I designed a decision-based workflow that captures the lifecycle of policy management, from initial AI use case submission to policy publication and enforcement.
DESIGN DECISIONS
STEP 1: Lowering the Barrier to Policy Creation by Introducing a Policy Composer
OPTION A - Generate with AI
Describe the use case in natural language
Receive a policy draft with suggested rules, conditions, and actions.
Refine and customize the generated policy.
OPTION B - Start from Scratch
Build a policy manually for highly specific or advanced governance scenarios.
Retain full control over every rule and workflow.
STEP 2: Making Policies Testable Before They Go Live
FAILURE STATE
What happens when a policy contains missing logic, disconnected rules, or incomplete decision paths?
RECOVERY FROM FAILURE STATE
If this policy fails to execute because of missing or broken logic, how can the system help users fix it?
IMPACT
The success was measured by how effectively users could create, validate, and confidently deploy policies.
92%
task completion rate for creating and publishing a policy
8/8
participants felt confident publishing a policy after running simulations
88%
users successfully recovered from policy failures using AI recommendations
LEARNINGS & TAKEAWAYS
How this project has helped me grow as a designer?
Trust Requires Explainability
Trade-Offs Drive Great Design
