← All work
Service Design → UX Design · Enterprise CRM · 2025–2026

Transforming Sales Workflows Through Service Design & AI

Enterprise Technology Platform (Confidential)

A two-phase engagement: first as a service designer shaping the AI-enabled future state, then re-engaged as a UX designer to execute and standardize the product.

Service DesignAI WorkflowsUX DesignSLDSEnterprise CRM
Enterprise CRM unified account plan interface
Phase 01
Service Design
Workshops · Current & future state · AI journey map
Phase 02
UX Design: One CRM
SLDS patterns · Design system · Product execution
Role
Lead Experience Designer
Timeline
Apr 2025 – Feb 2026
Team
Product, Engineering, Sales Ops
Client
Enterprise Technology (Confidential)
01
Phase 01
Service Design
Workshops, current & future-state requirements, AI journey mapping
Phase 1 · Discovery

From opportunity to
system-level insight.

We entered the engagement exploring how AI could accelerate and simplify sales workflows. What surfaced as a technology opportunity quickly revealed something deeper.

Across conversations with Sales, Ops, and Engineering, it became clear the challenge was not just about tools — it was about how teams worked. The focus shifted from introducing new technology to rethinking the system itself.

No shared vision for how sales teams should operate
Work fragmented across disconnected tools with no clear handoffs
AI capabilities existed but lacked ownership or application
Teams solving similar problems independently, with no system-level view
No end-to-end view of the sales lifecycle
Our Approach
Current-state workshops
3 cross-functional sessions
Stakeholder interviews
Sales, Ops, Engineering, IT
Process walkthroughs
Live tool and workflow observation
Future-state co-design
With Sales and AI SMEs
AI capability mapping
Where automation adds real value
Key Shift

What started as a technology conversation evolved into defining a new operating model for sales. We introduced the LAER model (Land, Adopt, Expand, Renew) as a way to structure how value is created across the lifecycle, aligning teams, workflows, and decision-making.

Primary Output
An interactive, end-to-end AI-enabled journey map grounded in the LAER model, bringing the future way of working to life
Artifact · End-to-End Journey Map

The future state, made visible.

Working with Sales and AI SMEs, we translated the new operating model into a concrete, navigable artifact. The map shows how value is realised at every step across the LAER lifecycle, with AI-enabled touchpoints mapped to actor, interaction type, and system ownership. It operates at two levels: L1 covers the full commercial cycle end-to-end, L2 breaks each stage into its sub-processes and handoffs.

L1 · End-to-End Journey
Journey in Action: Realizing Customer Value at Every Step. L1 overview across Land, Adopt, Expand, and Renew

The full L1 journey across Land, Adopt, Expand, and Renew, realizing customer value at every step. Each stage has AI-enabled touchpoints mapped to actor, interaction type, and system ownership.

L2 · Sub-Process Detail
L2 drill-down of the Adopt stage, showing sub-processes, actors, and interaction types

Clicking into an L1 stage expands the L2 layer: individual sub-processes with actor ownership, step type, and interaction notes. Shown here: the Adopt stage. Used as a shared alignment tool across Sales, Product, and Engineering.

Artifact · End-to-End Journey Map

The Future State, Mapped

The journey map spans the full customer lifecycle across four stages: Land, Adopt, Expand, and Renew. Select any L1 stage below to expand the L2 sub-process detail.

Land
Close the initial opportunity and set the foundation for a successful customer relationship.
2 sub-processes
01
Finalize
Sales Team
Confirm scope, terms, and stakeholder alignment before handoff
Manual
02
Prepare to Launch
Cross-functional
Coordinate internal teams and set up the account for day one
ManualAI-enabled
Key
Manual
System
AI
Approval
AI-enabled

The journey map worked. The client came back.

The service design phase established a clear future-state direction. That clarity led to a re-engagement as a hands-on UX designer, embedded with the internal team to build it.

Phase 1 output
Interactive AI journey map
Client decision
Re-engaged for execution
New role
Embedded UX Designer
Mandate
Standardize and accelerate product design
02
Phase 02
UX Design: One CRM
Embedded with the internal team to standardize, accelerate, and lead design
Phase 2 · The Problem

Developers were designing.
Design was cleaning up.

When I joined the internal design team, the process was broken in a specific way: developers were building screens and flows based on their own interpretation of requirements, then handing designs to UX for a redesign pass.

This created a cycle of churn. By the time design got involved, patterns had already been set in code. Redesigning them meant developer rework, schedule pressure, and friction that gradually pushed design further to the edges.

The result was an inconsistent product: different patterns solving the same problem, no shared vocabulary between teams, UX treated as a polish step rather than a foundational input.

The Churn Cycle
01
Requirements land
Dev and Design receive at the same time
02
Dev interprets & builds
UI decisions made in code without UX input
03
Design reviews post-build
Inconsistencies flagged after implementation
04
Rework requested
Dev revisits work already shipped to QA
05
Patterns diverge
Each cycle adds inconsistency to the system
Phase 2 · Response

Design at the front, not the end.

The goal was to invert the flow. Design provides the vision and the pattern, engineering builds to it. Four things made that possible.

Get ahead of the build

Established a practice of designing 1–2 sprints ahead of engineering. Design decisions were made in Figma, not in code.

Build the pattern library

Established a base design pattern library from SLDS, then adapted each component to the specific flows and interactions of the new CRM. Not just a theme. A system.

Standardize across the product

Audited existing screens and catalogued inconsistencies. Defined a single pattern for each repeated interaction: navigation, forms, data tables, empty states.

Make design a blocker (the good kind)

Worked with product to require design sign-off before any new component entered development. Design became a gate, not a review.

Phase 2 · The Product

One place. Full context. AI throughout.

The Unified Account Plan connected the entire CRM into a single structured experience. An account executive with no prior knowledge of an account could open it and immediately understand where they are, what matters, and what to do next. AI is embedded at every layer, not as a feature added on, but as the thing that makes the experience work at all.

01

Discovery and strategic priorities

The landing view surfaces an AI-generated account summary with Pillar Snapshots across Land, Adopt, Expand, and Renew, giving reps immediate context on where the account stands across the full lifecycle. From there, AI surfaces quick-start priority suggestions grounded in live market signals. No blank canvas or manual research required.

AI Summary shows account health and pillar snapshots across all LAER stages
Quick Start Priorities generated from market signal analysis
Source citations: news articles, customer calls, market intel
AI-drafted priority descriptions ready to refine and save
Discovery and strategic priorities screenshot
1 / 3
02

Technical initiatives and milestones

For each priority, reps define technical initiatives. AI drafts the full technical overview — including a technology profile and executive impact statement — pulling from Active Sources surfaced on the right. Initiatives are tracked with milestones, owners, and priority levels once saved.

AI suggestion chips for initiative titles based on priority context
Full technical overview and technology profile auto-generated
Active Sources panel grounds content in real account signals
Milestone tracking with owner assignment and priority levels
Technical initiatives and milestones screenshot
2 / 4
03

Outcomes and metrics

The final layer ties initiatives to measurable outcomes. AI drafts the outcome description and suggests how to describe the expected impact. Outcomes link back to their parent initiatives, giving the account executive a complete picture from strategy to delivery.

AI-drafted outcome descriptions tied to initiative context
Active Sources referenced throughout for grounded suggestions
Quick start prompts to build out additional outcomes
Outcomes linked back to parent initiatives for full traceability
Outcomes and metrics screenshot
1 / 2
Phase 2 · Pattern Library

SLDS, adapted. Not just applied.

Salesforce Lightning Design System provided the foundation. But a foundation is not a product. Every component needed to be adapted for the specific flows, density, and AI interactions of the new CRM.

Artifact · UX Framework Spec
UX framework spec showing Search Module and Table Module structure with annotated component rows

The framework spec defined the structural rules for two core modules used across every page: the Search Module (query row, refinement row, action row) and the Table Module (view controls, column hierarchy, pagination). These became the shared language between design and engineering.

Artifact · Search Patterns Across Contexts
Opportunity Unified View search with refinement fields
Screens
Opportunity Search

The same search module applied consistently across Opportunity Search, Account Search, and Quote View. Different data contexts, same structural pattern. Annotations visible in the specs were shared directly with engineering to enforce spacing and min-width rules.

Navigation
Global nav
Tab bars
Breadcrumbs
Side panels
Data Display
Record detail
List views
Related lists
Inline edit
Forms & Input
Field groups
Picklists
Lookup fields
Validation
AI Surface
Insight cards
Action prompts
Confidence labels
Dismiss/act
Status & State
Loading states
Empty states
Error handling
Success flows

"The pattern library wasn't just documentation. It was the agreement between design and engineering on how the product should behave."

Craft

Key Decisions

The calls that shaped how the product was built, and how design held its ground.

Vision vs Execution

Design the workflow before the screen

Every major feature was mapped as a user flow before any UI work began. This kept alignment ahead of the build and prevented scope creep in components.

Speed vs Consistency

Standardize first, optimize later

Resisted pressure to customize individual screens. Getting consistent patterns in place, even imperfect ones, was more valuable than bespoke solutions that couldn't scale.

Trust vs Control

AI that shows its reasoning

Every AI surface was designed with a visible reason for the suggestion. Trust came from transparency, not accuracy alone.

Influence vs Authority

Alignment as the output

In the absence of formal design authority, ran structured reviews and decision logs. Every major call was visible, auditable, and reasoned. That gave design decisions staying power.

Impact

What it moved

MVP
Contributed to launch
On time, on scope
SLDS
Pattern library established
Adapted to CRM flows and AI surfaces
Workflow complexity
Consolidated multi-step flows
Design-to-dev velocity
Design ahead of build by 1–2 sprints
Re-engaged by client
Service Design → Product Execution
1
Consistent system
Across all product teams
Reflection

What I'd carry forward

01

Systems thinking is critical in enterprise design. The interface is rarely the constraint. The organizational process usually is.

02

Getting ahead of the build is a strategy, not a luxury. Design has the most leverage before anything is coded, not after.

03

Alignment is often the highest-impact deliverable. A shared pattern is worth more than a polished one-off.

Due to the confidential nature of this work, details and visuals have been adapted. Additional materials available upon request.