Strategy Lead

Strategic GenAI Chatbot Design for Fortune 10

Accelerating resolution and reducing live agent escalations through intelligent automation

A Fortune 10 cloud provider serving millions of prospects and customers required intelligent self-service solution at scale. I led the strategic direction and design of GenAI Chatbot 2.0—an intuitive experience that accelerated product discovery, elevated information access, and enabled users to find answers quickly. The chatbot delivered empathetic conversational flows with seamless escalation to specialists when needed—setting a new benchmark for support usability at scale and enabling faster resolution with improved customer satisfaction.

Usability TestingCompetitive AnalysisUX WritingGen AI Chatbot Experience StrategyConversation DesignUI & Interaction DesignData SynthesisWireframingUse Case PrioritizationEnterprise Experience DesignDecision ArchitecturePrototyping & Testing

Challenge

Prospects and existing customers struggled to find relevant information through the chatbot, often abandoning tasks or escalating unnecessarily. Without a clear understanding of user intent, discovery patterns, or common support journeys, the team lacked the insight needed to design a seamless, scalable self-service experience that could adapt across segments and entry points.

Problem Statement

How might we design an intelligent self-service chatbot experience that empowers agents to focus on complex, high-impact interactions?

Role & Responsibility

Role & Responsibility

UX Strategy for Chatbot Experience
Scope Definition
Conversational UX Strategy
Feature Prioritization
Data Analysis & Synthesis
Design Mentorship
UI & Interaction Patterns

Deliverables

Audited competitors chatbot experience
Analyzed automation opportunities
Established decision architecture for human handoff
Mapped how, where, when, why of escalation points
Designed conversation flows
Prototyped key conversational interactions
Ran iterative user testing on self-service scenarios
Developed modular conversation components
Authored UX writing for conversational moments

User Research

Competitive Audit

Conducted a competitive audit of 15 cloud industry companies to benchmark self-service capabilities, escalation models, and conversational best practices; synthesized insights from SMB-focused user interviews and analyzed live support transcripts to identify patterns in human-to-human interactions.

UX Strategy

GenAI Chatbot Experience

Defined project scope and surfaced key experience moments by mapping the end-to-end chatbot journey—from initial engagement to termination. Identified high-impact intervention points and established a strategic foundation to streamline interactions, refine decision logic,

How might we design an intelligent self-service chatbot to enable agents to focus on more complex, high-impact interactions?

Intuitive Welcome Experience

Develop prompts that align with users’ mental model of learning in the default state to maximize user engagement from the outset.

Empathetic Conversational Flows

Create a natural conversation framework resembling human-to-human interactions, specifically through transitional sequences.

Seamless Human Escalation

Identify friction points in the current flow and define key moments of minimimum barriers for a smoother handoff.

Discreet Feedback Collection

Prompt users for input without disrupting their flow by timing feedback requests around natural pauses and using minimal, unobtrusive UI.

UX Design

Intuitive Welcome Experience

Designed mulitple concepts of the chatbot welcome experience that helps users quickly articulate their needs, engage seamlessly with the chatbot, and navigate its full capabilities with confidence and ease.

How might we optimize the Welcome Experience to expedite the user task?

Relevant Prompts

Intuitive UI

Frictionless Flow

Empathetic Conversation Flows

Identified recurring escalation scenarios and modeled the AI-to-human interaction as a system of states and decision points rather than a single flow. The framework defined how intent, context, and emotional continuity are preserved across retries, failures, and handoffs—establishing consistent escalation behavior that could scale across use cases and teams.

How might AI-to-human escalation behave across intent, failure, and recovery states?

Intent Modeling

Escalation Logic

State Continuity

Scenario 1:

User initiates human escalation and the system guides the user to appropriate support.

“I’m satisfied with the chatbot’s responses and need human support.”

Scenario 2:

System initiates human escalation upon detecting user dissatisfaction or an unanswerable query (e.g., pricing).

“I’m NOT satisfied with the chatbot’s responses and need human support.”

Scenario 3:

User initiates human escalation, and the system first attempts to assist before directing them to appropriate support.

“I’ve not interacted with the chatbot and I need human support.”

Seamless Human Escalation

Developed a simplified escalation flow from GenAI to human agent that eliminates redundant input requests and clearly guides users through key steps—offering flexible handoff options such as scheduling a call, and enabling continued access to GenAI assistance throughout the transition.

How might we design intelligent and empathetic escalation flows that guide users seamlessly to human support?

Dual Access

Contextual handoff

Clear Transition

Discreet Feedback Collection

Developed a strategic framework for capturing feedback at both the conversation and response levels, designed to minimize friction and preserve the integrity of the user’s task flow. Defined 5 pivotal moments for optimal engagement, grounded in when and why users are most likely to respond. Designed contextually integrated prompts that align with where and how users interact, enabling feedback collection to occur naturally within the experience.

How might we unobtrusively gather real-time insights into users’ experiences?

Context-Aware

In-Context

Natural

Hi-Fi Design

(Credit: Provided design direction to the Design team on UI and interaction design.)

Welcome Experience

Conversation History Flow

Chat Flow w/ both GenAl and a Live Agent

Feedback Collection — Chat w/ Live Agent

Success Metrics

Impact at Two Levels

When engaged through an agency, I effectively have two clients—the agency that hires me and the client they serve. In this engagement, success meant:

Stabilizing a stalled project post-vendor exit and delivering the GenAI chatbot on an abbreviated timeline
Restoring Instrument’s client relationship, expanding their AI delivery credibility, and mentoring the internal team

Impact to Client: Fortune 10

Agent Offload: Cut live-agent escalations by 30% across platform serving millions of customers, allowing agents to focus on high-value inquiries.

Task Completion: Boosted chatbot task completion by 25%, accelerating resolution and improving user satisfaction.

Scalable Patterns: Standardized interaction models across 3 business units, enabling faster deployment of future AI features.

Timeline Recovery: Delivered chatbot in 2 months, recovering 6 months of stalled progress to meet client deadline

Impact to Agency: Instrument

Trust Recovery: Restored Instrument’s credibility with Fortune 10 client, securing contract extension for broader client initiative.

Capability Expansion: Proved Instrument’s ability to lead complex UX and GenAI work—beyond its core brand offering.

Team Enablement: Upskilled 5-person design team, enabling independent delivery on subsequent AI projects.