Case study · AI + automation
Turning customer conversations into automated business workflows.
Small businesses often rely on manual processes to answer questions, qualify leads, capture customer information, and coordinate follow-up. I designed and prototyped conversational AI experiences that provided 24/7 customer service and lead intake while connecting conversations to automated workflows and business systems.
Challenge
Reduce repetitive customer-service work without creating a rigid or frustrating automated experience.
My role
Product Designer / AI Experience Designer
Led customer discovery, conversation design, workflow design, prototyping, validation, and implementation across multiple client use cases.
Impact
24/7 service
Customer service + lead intake beyond business hours
The challenge
Small businesses were spending time answering recurring questions, qualifying leads, collecting customer information, and coordinating follow-up manually. When someone wasn't available, or a customer communicated in another language, potential inquiries could be harder to serve immediately.
The opportunity was not simply to add a chatbot. The experience needed to understand conversational requests, identify what the customer was trying to accomplish, collect the right information, and connect the conversation to the business process behind it.
My role
I worked directly with business owners to understand customer needs and operational workflows, then translated those requirements into conversational experiences, automation logic, and working prototypes. I designed both sides of the experience: what the customer interacted with and what happened behind the conversation.
Key decisions
01
Rather than starting with the technology, we identified the customer questions, requests, and service interactions that created repetitive work for the business. That focused conversational automation where it could make the interaction easier for customers while reducing unnecessary manual effort.
02
The assistant needed to do more than answer questions. I designed conversations that could gather the information required to move a request forward and hand it to the appropriate business process — a bridge between customer intent and action.
03
The conversational interface was only the customer-facing layer. I designed the broader experience around the workflows, business rules, data, and follow-through behind the conversation, so customer interactions connected to real operational processes and could be adapted per business.
The solution
The resulting experiences combined conversational assistance with workflow automation behind the scenes. Customers could ask questions, provide information, or request help through a natural conversation, while the experience connected those interactions to the business processes needed to move the request forward.
01 · Customer service
Answer common questions across languages and guide customers toward the information, service, or next step that fits what they asked for.
02 · Private jet booking concierge
Guided sales and service: help travelers explore relevant charter options and collect what the operator needs to follow up on a trip request.
03 · AI intake assistant for law firms
Client intake: set expectations up front, triage urgency, capture case details, and route toward a consultation — without giving legal advice.
04 · Adaptable to different industries
The same conversation-to-workflow model applied to a boutique travel client, using its own knowledge, resort portfolio, and concierge tone of voice.
Impact
24/7 service
Questions and lead intake
Customer service and lead intake can continue beyond normal business hours
Multilingual
Requests across languages
Supports conversational customer interactions across languages and routes requests by intent
Automated
Conversation to workflow
Customer conversations can trigger intake, CRM updates, summaries, routing, documentation, and follow-up
Reusable
Common agent framework
A common agent framework can be adapted across different businesses and workflows
We didn't just design a chatbot. We connected the conversation to the business process behind it.