Prime Squad - AI Voice Agent for Real Estate
A real estate company focused on helping clients navigate property opportunities while managing leads, communication, and follow-up throughout the sales process.
Client:
The Tycoon Realty
Category:
Workflow Design, UI & UX Design, Dashboard Design
My Role:
Lead Product Designer
Live Preview
Live Preview
RAITECH AI developed an AI voice automation solution for Tycoon, a real estate company managing leads, client communication, and follow-up throughout the sales process.
The solution helps Tycoon automate lead responses, qualification, and follow-up while connecting the AI agent with their existing workflows.
I focused on understanding the existing workflow, identifying opportunities for AI automation, and designing a connected experience across the AI agent, users, workflows, and supporting systems.
My Role & Responsibilities
As Lead Product Designer, I led the product design from discovery and ideation through UX, UI, prototyping, and delivery.
Mapped end-to-end workflows
Defined the AI agent persona and conversational behavior, including tone, scripts, and response flows
Designed conversational UX for AI voice
Designed human-in-loop experiences
Designed automation workflows
Designed the dashboard and surfaced key metrics
Applied systems thinking to connect users, AI agents, workflows, data, and supporting systems into a cohesive experience.
Why are we doing this?
Tycoon Real Estate wanted to modernize its workflow by introducing an AI voice agent to automate repetitive tasks, respond to leads faster,and support agents throughout the lead management process.
The goal was to integrate AI into the existing workflow to reduce manual work, improve response time, and help real estate teams focus on higher-value conversations.
Research & Discovery
Before designing the AI experience, I first focused on understanding how the existing lead process worked, where agents spent the most time, and where manual work was creating friction. This helped identify the right opportunities for AI while keeping the human relationship at the center.
Interviewed stakeholders to understand how inquiries were handled
Mapped the current-state journey from inquiry to agent follow-up
Identified repetitive tasks, decision points, and manual handoffs
Identified opportunities where AI could reduce manual effort
Design principle: Automate the work, not the human relationship.
How Do We Solve it?
We translated the research findings into a more resilient AI experience one that can understand context, handle uncertainty, and know when to involve a human.
Instead of designing only for the ideal conversation, I mapped the key edge cases and defined clear behaviors for each scenario. The AI was designed to clarify when intent is uncertain, update the conversation context when needs change, escalate when a request is outside its capabilities, and allow human correction when something goes wrong.
This approach helped create an experience that feels helpful and controlled, while keeping the customer moving forward without unnecessary friction.
Edge Case Handling Workflow
The workflow visualizes how the AI evaluates each interaction and determines the appropriate response—from clarification and context updates to human handoff and correction.
Customer Input → AI Evaluation → AI Response → Outcome
Key design principles
Clarify, don’t assume — Ask when intent is unclear.
Adapt in real time — Update context as needs change.
Know the limits — Escalate when the AI cannot safely help.
Keep context — Preserve conversation history across handoffs.
Enable human recovery — Allow correction without restarting the journey.
Mia is a thoughtful, dependable, and highly collaborative Senior Product Designer. She brings clarity to complex problems, combines strong attention to detail with empathy for users, and consistently delivers high-quality, purpose-driven work. She works exceptionally well across teams and always brings a positive, thoughtful approach to collaboration. I would gladly work with Mia again."
Senior UX Leader
Leon Copeland
Related Projects
