Reputation GenAI & Automation Strategy

Augmenting Design Velocity: Embedding Agentic AI Workflows in Enterprise SaaS

How an enterprise UX organization integrated advanced prompt engineering, custom LLM agents, and design system automations to shield creators from rote asset production and optimize initial discovery phases.

Role

Director of UX / Principal Designer

The Mandate

Scale output quality without adding linear headcount dependencies.

Core Tech Stack

Gemini Suite (Gems, Notebook LLM, Workspace), AI Studio, Figma Make, Spark (ProductBoard), Zapier

Key Outcome

80% time savings due to AI tools used in PRD breakdowns and design reviews.

The Leader Challenge

As Director of UX at Reputation, I inherited a global portfolio spanning flagship sentiment management services and multiple complex individual product modules. The design team faced a dual challenge: managing aggressive feature development tracks across international engineering hubs, while simultaneously modernizing legacy interfaces to meet rigid WCAG 2.2 usability and accessibility standards, and fixing all existing usability issues across the platform.

Traditional scaling models dictate adding linear design and research headcount to match expanding engineering pipelines. To bypass this friction, I strategized an operational pivot: **up-skilling the existing team to utilize local and cloud-based AI workflows, turning designers from raw asset producers into strategic editors.

J. Kurt Iverson Portfolio Visual

The Three-Tier AI Integration Strategy

Rather than deploying superficial out-of-the-box AI features, I helped facilitate a multi-organization, governed, human-in-the-loop automation layer tailored directly to our internal SDLC processes.

1. Agentic Tools (Gemini Custom Gems and Gemini CLI)

We ingested massive repositories of qualitative user interviews from UX Lab, Pendo usage tracking data, UX guidelines, WCAG requirements and historical journey maps, training specialized custom **Gemini Gems. These internal AI agents act as dynamic focus groups. Designers can instantly upload a draft Product Requirement Document (PRD) and prompt the agent to stress-test the features against known user frustrations, compressing initial UX discovery timeframes. As a design supervisor, I have personally decreased my voice and tone/grammar reviews by 90%. Having created a voice and tone markdown file and hooked Figma MCP to Gemini CLI, I was able to prompt an analysis based on a URL and review issues in the background.

2. Rapid Functional Prototyping (Google AI Studio)

To accelerate early layout validation, the team utilized **Google AI Studio** pipelines. By feeding the system structurally validated platform frameworks, designers can construct multi-variant navigation wireframes dynamically from text-based structural outlines, letting the team test layout hypotheses in real time before diving into high-fidelity component tracking.

3. Prototype to Design File (AI Studio, Figma Make and Gemini Gems)

To decrease time between concept and baseline designs, we use AI Studio to flesh out initial functional flows. Product and Engineering have access to UX design patterns and can get from concept to stylistically aligned prototypes in minutes. Once PMs and UX agree on the approach, prototype screens are transferred to Figma Make and converted to our final styles. Quick voice and tone, WCAG and style checks are completed using Gems before final hand off to engineering. This completely bypassed hours of manual alignment work, allowing designers to allocate their cognitive energy entirely to solving structural flow logic, information hierarchy, and complex usability criteria.

Measurable Business Transformation

This was not an academic experiment in automation or efficiency; it drove rapid, concrete market results for the enterprise. By combining targeted AI automation with rigorous design refactoring, we cleared extensive user cognitive load across our highest-friction tools—such as the core enterprise Reviews and Survey Builder ecosystem. We are pragmatic about our use of AI as well. Like AI tools augmenting our own processes, we see AI as a highly focused assistant for our consumers rather than their replacement.

2 Months
80% AI Tool Use
WCAG 2.2
Strict Accessibility Compliance
Zero
Linear Headcount Dependencies

The compounding velocity gained from our AI workflows allow the design team to stay ahead of cross-functional engineering sprints. Within two months of implementation, the optimized product flows successfully transformed user engagement data, visibly improving platform perception among enterprise customers and our corporate board of directors. Additionally, we were able to increase design speed on 6 new product offerings for Transform 2026 (something that has never happened before).

Want to discuss embedding these operational frameworks into your design org?
← Dashboard