Genesis AI
Platform designed to help sales and operations teams efficiently create, manage, and present product configurations for enterprise clients.
The legacy CPQ was outdated, slow, and not aligned with the industry shift toward AI-assisted configuration. Users struggled with hidden complexity, disconnected admin logic, and workflows that demanded far more effort than necessary.
I redesigned the whole system by reimagining the system from the ground up — auditing every legacy flow, co-creating with stakeholders, and designing a modern, AI-assisted experience that balances automation with human control.
“How do you redesign a legacy enterprise configurator used by sales teams every day—without disrupting years of learned behavior, while introducing AI as the primary way of configuring complex products?”
Project Snapshot & Responsibilities
My responsibilities extended far beyond designing interfaces. I worked across the complete product lifecycle—from understanding business strategy and legacy workflows to collaborating with engineering on implementation.
My responsibilities included:
- Product discovery alongside CEO & leadership
- Auditing the legacy configurator
- Mapping End User & Admin user journeys
- Facilitating stakeholder workshops
- Collaborating with PMs & VP of Design
- Aligning solutions with dev feasibility
- Designing scalable interaction patterns
- Supporting implementation through design QA
Designing Two Interconnected Panels: Admin Side & User Side
A critical aspect of this project was designing two distinct yet synchronized panels that work in unison across the enterprise CPQ ecosystem.
System Setup, Catalog & Rule Engine
Designed for system administrators to define the foundation of the product configurator:
- • Creating & structuring product catalogs (eCatalog)
- • Defining Categories, Item Types & Options
- • Setting pricing formulas, discounts & dependencies
- • Configuring rule logic (inclusion/exclusion rules)
- • Managing cross-sell & recommendation rules
Sales Representative Configurator
Designed for sales teams to build customer quotes efficiently:
- • Conversational AI intent-driven configuration
- • Dynamic option comparison & selection cards
- • Real-time rule validation & conflict warnings
- • Fast 1-click quote assembly & export
- • Polished interface ready for live client demos
More importantly, I became responsible for something less visible: creating clarity inside a highly complex enterprise product while ensuring every design decision balanced user needs, technical constraints, and business objectives across both Admin and End-User panels.
Working on Genesis AI
Current photograph from Kochi workation while architecting Genesis AI.
Transforming Enterprise CPQ Workflows
Genesis AI wasn't simply a redesign project. It represented a shift in how enterprise sales teams interact with CPQ (Configure, Price, Quote) systems. Instead of navigating deeply nested configuration screens, sales representatives could describe customer requirements conversationally while AI generated product configurations in real time. The redesign focused on reducing complexity without reducing flexibility.
Project Highlights
Hero Mockup
Large full-width configurator interface showcasing the AI assistant and intent workspace.
The Opportunity & The Challenge
The Opportunity
Traditional CPQ systems are incredibly powerful—and notoriously difficult to use. Sales reps spend more time configuring products than selling. As product complexity grows, so does the number of rules, dependencies, and validations.
The opportunity was to fundamentally rethink how enterprise users interact with complex configuration systems by exploring AI as the primary interface while preserving precision.
The Challenge
Enterprise configuration isn't like designing a simple form. Every product contains hundreds of combinations, dependencies, and business rules before a quote can be generated.
We needed to balance two conflicting goals: introduce a completely new AI-assisted workflow without disrupting existing user expectations and learned habits.
Before & After
Side-by-side comparison: Legacy Configurator beside redesigned AI-assisted configurator.
Aligning on the Vision
Every successful product begins with alignment. Our journey started with a discussion with the CEO—not to review designs, but to establish a shared understanding of what success should look like. Rather than discussing colors, layouts, or components, the conversation focused entirely on business outcomes.
The Vision Was Clear:
- •Sales reps should spend less time configuring and more time building relationships.
- •AI should reduce manual effort instead of adding another layer of complexity.
- •The product itself should become the live demo presentation.
CEO Workshop
Mapping key strategic outcomes and business vision with leadership.
Key Themes
My Design Mindset
Rather than jumping directly into interface ideas, I wanted to understand the reasons behind every business objective. Throughout the project I continuously asked questions like:
Understanding the Legacy Product
Before proposing solutions, I needed to understand how the existing configurator actually worked. Rather than reviewing a handful of screens, I imported hundreds of legacy flows into Miro and treated the process like a forensic investigation. The objective wasn't to identify visual inconsistencies—it was to uncover structural problems that stakeholders had gradually become accustomed to over years of usage.
Legacy Configurator Audit
Large Miro board mapping hundreds of legacy flows, parameter grids, and friction points.
The Audit Focused On:
- •Where users hesitated
- •Where cognitive overload occurred
- •Which workflows repeatedly caused confusion
- •Which interaction patterns broke expectations
- •Where information hierarchy broke down
- •Which workflows created unnecessary manual effort
Stakeholder Workshops & Key Insights
The configurator supported multiple business teams. To ensure the redesign reflected real operational needs, we conducted collaborative workshops with stakeholders across Sales, Operations, Product, and Administration. Using the Miro audit as a shared workspace, we collectively reviewed workflows and highlighted critical features, usability issues, and opportunities for AI assistance.
Stakeholder Workshop
Cross-functional alignment sessions in Miro with Sales, Ops, and Admin leadership.
Listening Beyond Requests
Stakeholders naturally described symptoms such as “The configurator feels slow” or “Users keep making mistakes.” Instead of accepting statements at face value, I asked: “What are users trying to achieve at this point?” and “What forces users to stop and think?” to uncover root workflow problems.
4 Core Discovery Insights
1. Configuration was overwhelming
Users weren't struggling because products were inherently complicated—they struggled because the interface exposed too much information simultaneously.
2. AI should simplify—not replace—decisions
Enterprise users still wanted visibility and control. AI needed to become a collaborative assistant rather than an autonomous decision maker.
3. Context frequently disappeared
As users navigated deeper into configuration, they lost awareness of where they were and what product they were currently modifying.
4. Flexibility introduced inconsistency
Administrator-controlled layouts created dozens of unpredictable interface combinations, making the product difficult to scale.
5 Guiding Design Principles
1. Make AI feel like a teammate
AI should reduce effort while preserving transparency and user confidence.
2. Preserve familiar mental models
Innovation shouldn't require users to relearn years of product knowledge.
3. Reveal complexity progressively
Users should encounter information only when it becomes relevant.
4. Design for variability
Interfaces should remain stable regardless of admin configs or product complexity.
5. Prioritize scalability
Every component should support future products, industries, and models without redesign.
Design Principles Illustration
Visual architectural principles mapping progressive disclosure and AI teammate dynamics.
From Discovery to Initial Iterations
Only after understanding the business vision, auditing the legacy system, and aligning with stakeholders did we begin exploring interface directions. Our first iterations explored how AI, navigation, hierarchy, and component architecture could work together as one cohesive experience.
First Figma Exploration
Early high-fidelity layout iteration mapping AI panel side-by-side with parameter grids.
Designing the Solution & The AI Panel
Discovery gave us a clear understanding of the problems. The next step was designing a configuration experience that reduced decision-making without reducing flexibility.
Final Configurator Overview
Full-screen mockup displaying the AI left panel, collapsing header, sticky category headers, and responsive option cards.
The AI Panel: Primary Gateway to Configuration
The AI panel became the centerpiece of the redesign because it reversed the traditional workflow: instead of asking users to configure products manually step-by-step, they describe needs conversationally while AI generates the initial configuration. Manual interaction became a refinement step.
AI Panel Interaction
Animation showing AI expanding into the workspace, evaluating parameters, and generating a validated configuration.
Why the AI Lives on the Left Side
Users already associate the left side of the screen with navigation, conversation, and workflow guidance (ChatGPT, GitHub Copilot). Placing AI on the left reduced the learning curve and communicated its primary role.
What the AI Panel Enables:
- Describe requirements conversationally
- Generate product configs automatically
- Resolve conflicts through contextual prompts
- Understand underlying configuration logic
- Spend more time talking with customers
- Eliminate repetitive manual parameter entry
Designing Persistent Context & Navigation
Enterprise configurations often contain hundreds of options. As users progressed through the workflow, important context gradually disappeared. Several key interaction decisions focused on preserving orientation.
Header Redesign: Collapsing Header
As users scroll deeper, the hero header smoothly transitions into a compact sticky bar preserving product image, name, primary actions, and key identifiers—maintaining orientation while maximizing workspace.
Collapsing Header
Micro-interaction video showing header compressing into compact persistent navigation bar.
Sticky Item Type Headers
Each Item Type header remains docked to the top while its associated options are in view, providing a persistent visual anchor for section context during long configuration sessions.
Sticky Category Prototype
Prototype showing section headers pinning dynamically as the workspace scrolls.
Information Hierarchy & Option Card Architecture
3-Level Hierarchy
Hierarchy Diagram
Visual diagram mapping Item Category → Item Type → Configuration Options structure.
Rethinking the Option Card System
The option card was the most technically demanding component. It needed to adapt gracefully to administrator settings (enabling/disabling images, pricing, quantity controls, or badges) without layout breakdown.
Legacy Card vs Redesigned Card
Side-by-side comparison of cluttered legacy option cards vs streamlined modular card architecture.
Card Variant Matrix
Exhaustive design token matrix mapping all administrator-enabled layout states.
1-Up • 2-Up • 3-Up Card Variants
Responsive grid layouts adapting across Full Width (1-Up), Two Column (2-Up), and Three Column (3-Up).
Business Alignment & Technical Constraints
Business Priorities
- • Sales Teams: Faster configuration across industries.
- • Administrators: Fewer errors & clearer validation rules.
- • Product Management: AI-first market readiness.
- • Leadership: Modern enterprise CPQ demo platform.
Technical Realities
- • Legacy Data Structure: Backend nested attribute limits.
- • Rules Engine: Sequential validation processing constraints.
- • Navigation Arch: Preserving core structural frameworks.
- • AI Engine Sync: Real-time conversational sync with CPQ logic.
Technical Constraints & System Architecture
System architectural blueprint aligning design tokens with backend rules engine constraints.
Design Review Process
Continuous cross-functional review cycles between Leadership, PM, Engineering, and Design.
Outcomes & Reflection
Outcomes
The redesign established a stronger foundation for the next generation of the Genesis configurator. By combining AI-assisted workflows with clearer information hierarchy and scalable interaction patterns, the experience became significantly easier to navigate while remaining compatible with existing enterprise workflows.
Final Product Showcase
Complete Genesis AI Configurator design showcase across End User and Admin panels.
What I'm Most Proud Of
Looking back, the achievement I'm most proud of isn't a single interface or interaction—it's the process. Despite product complexity and delivery pace, collaboration remained consistent across leadership, product management, design, and engineering. The result wasn't simply a redesigned configurator—it was a shared understanding of what the future of enterprise configuration could look like.
Reflection
Working on Genesis AI fundamentally changed how I think about enterprise software. At the beginning of the project, I viewed configurators primarily as interfaces for selecting options. Over time, I realized the real challenge wasn't configuration itself—it was helping users make confident decisions inside an environment filled with complexity.
Rather than asking “How can we redesign this screen?”, I began asking “How can we reduce the cognitive effort required to reach the right decision?” That question continues to shape how I approach enterprise product design today.
Team Photo / Closing Hero
Closing project celebration & team snapshot.
Thank you for taking the time to explore this case study. If you'd like to discuss enterprise UX, AI-assisted workflows, design systems, or product strategy, I'd love to connect.