Service Blueprinting AI-Human Boundaries in Onboarding Systems
Service Blueprinting AI-Human Boundaries in Onboarding Systems
The bottleneck wasn't capacity. It was the absence of a service blueprint.
CONTEXT
CONTEXT
An award-winning interior design studio was losing 2–3 hours of senior designer time daily to repetitive onboarding inquiries.
The instinct was to automate, the brief: Design a chatbot experience. The real problem was more structural: there was no defined service blueprint to distinguish what required designer judgment from what didn’t. Without that clarity, every interaction defaulted to the designer. Scaling required designing the system boundaries, not just introducing a chatbot.
An award-winning interior design studio was losing 2–3 hours of senior designer time daily to repetitive onboarding inquiries. The instinct was to automate, the brief: Design a chatbot experience. The real problem was more structural: there was no defined service blueprint to distinguish what required designer judgment from what didn’t. Without that clarity, every interaction defaulted to the designer. Scaling required designing the system boundaries, not just introducing a chatbot.
MY ROLE
MY ROLE
Led service blueprinting and end-to-end design of an AI-powered onboarding system
Led service blueprinting and end-to-end design of an AI-powered onboarding system
Defined intent models, conversation architecture, and escalation logic by mapping AI–human service boundaries, translating them into CRM workflows and operational systems.
Defined intent models, conversation architecture, and escalation logic by mapping AI–human service boundaries, translating them into CRM workflows and operational systems.
APPROACH
APPROACH
Analyzed 20+ onboarding call transcripts to identify where designer expertise was genuinely required vs. used as fallback
Analyzed 20+ onboarding call transcripts to identify where designer expertise was genuinely required vs. used as fallback
1:1 Interviews with designers to surface friction in current workflows
1:1 Interviews with designers to surface friction in current workflows
Client surveys to map anxieties and drop-off points.
Client surveys to map anxieties and drop-off points.
Intent-based personas and an intent modeling framework to understand user motivations
Intent-based personas and an intent modeling framework to understand user motivations
Service blueprint mapping across AI + human touchpoints, defining frontstage, backstage, and decision layers.
Service blueprint mapping across AI + human touchpoints, defining frontstage, backstage, and decision layers.
Progressive disclosure conversation design to operationalize the blueprint in the chatbot experience
Progressive disclosure conversation design to operationalize the blueprint in the chatbot experience
RESEARCH ARTIFACTS
RESEARCH ARTIFACTS
Chatbot Failure Analysis: Identifies where the system defaulted to designers due to lack of triage logic.
Chatbot Failure Analysis: Identifies where the system defaulted to designers due to lack of triage logic.
Intent-Based Personas: Reframes users based on intent behind queries, not just demographics.
Intent-Based Personas: Reframes users based on intent behind queries, not just demographics.
Intent Modeling Framework: Structures user intent into decision logic that informs AI vs. human handling.
Intent Modeling Framework: Structures user intent into decision logic that informs AI vs. human handling.
KEY INSIGHT
KEY INSIGHT
Designers weren’t overloaded because there was too much work. They were overloaded because there was no service blueprint defining where expertise belonged.
Designers weren’t overloaded because there was too much work. They were overloaded because there was no service blueprint defining where expertise belonged.
Every question, regardless of complexity, escalated to the same place.
Every question, regardless of complexity, escalated to the same place.
The system lacked a triage layer to distribute interactions appropriately.
The system lacked a triage layer to distribute interactions appropriately.
Structuring onboarding around user intent + service boundaries enabled expertise to be preserved for high-value decisions
Structuring onboarding around user intent + service boundaries enabled expertise to be preserved for high-value decisions
SERVICE BLUEPRINT- The System That Fixed It
SERVICE BLUEPRINT- The System That Fixed It
The blueprint designed defines how interactions are distributed across AI and human layers, introducing a triage system that preserves designer expertise.
The blueprint designed defines how interactions are distributed across AI and human layers, introducing a triage system that preserves designer expertise.
Defines how interactions are distributed across AI and human layers, and where designer expertise is activated.
Defines how interactions are distributed across AI and human layers, and where designer expertise is activated.
This became the foundation for all downstream decisions, AI behavior, escalation logic, and workflow design.
This became the foundation for all downstream decisions, AI behavior, escalation logic, and workflow design.
Multi-Turn Conversation Architecture: Translates the service blueprint into staged interactions and controlled escalation paths.
Multi-Turn Conversation Architecture: Translates the service blueprint into staged interactions and controlled escalation paths.
OUTCOME
OUTCOME
Service blueprinting reframed the chatbot from a support tool into core service infrastructure. It didn’t just improve the system, it made scaling possible by structuring how expertise is accessed.
Service blueprinting reframed the chatbot from a support tool into core service infrastructure. It didn’t just improve the system, it made scaling possible by structuring how expertise is accessed.
10–15 hours/week reclaimed per designer
85% client satisfaction
40% inquiries resolved without escalation
1-week faster project kickoff
10–15 hours/week reclaimed per designer
85% client satisfaction
40% inquiries resolved without escalation
1-week faster project kickoff
Open to new roles- Seattle & Remote
Looking for research and strategy roles where the insight
actually shapes what gets built, not just validates it after the fact.
Nitya Jois | Seattle WA
Open to new roles- Seattle & Remote
Looking for research and strategy roles where the insight actually shapes what gets built, not just validates it after the fact.
Nitya Jois | Seattle WA
Open to new roles- Seattle & Remote
Looking for research and strategy roles where the insight
actually shapes what gets built, not just validates it after the fact.
Nitya Jois | Seattle WA