When it comes to concrete applications of AI in your UX work, Dscout MCP is streamlining laborious tasks that used to take hours of intense manual work. 

Knowing in the abstract about these features is one thing. Knowing how to actually apply them to everyday use cases is another. 

That’s why we pulled together an AI Show and Tell panel to demonstrate exactly how these features can transform your workflow and bring you into the next era of UX research. 

The panelists of included:

• Michael Winnick, Founder and CEO of Dscout 
• Charley Olman, Senior Research Advisor at Dscout
• Laura Oxenfeld, UXR Leader at SiriusXM

Many of the use cases are easier to watch than read, so we attached some key sections below. If you'd rather watch the whole webinar, check it out here.

Before diving in, what is an MCP?

Explaining technical concepts can get tricky, but think of an MCP (Model Context Protocol) as a conversational, easy-to-use version of an API (Application Programming Interface).

While traditional APIs connect software systems using rigid code commands, an MCP provides AI models with context and descriptions about what exists inside a tool. 

This allows users to connect different applications and retrieve or move data using natural, everyday language—like asking the AI to pull "that project I worked on yesterday about fans"—instead of typing out complex database queries or manual file exports.

How Dscout MCP works

Dscout MCP directly links AI workspaces (such as Claude or Microsoft Copilot) to the research data stored inside Dscout. Built with platform-specific knowledge and guardrails, it understands Dscout’s unique question types, project history, and account structures. 

This enables researchers to ask AI assistants to pull project data, analyze transcripts, or draft launchable research projects straight from their primary workspaces without needing manual CSV exports or constant platform switching.

While MCPs have countless use cases, during the webinar, Laura, Charley, and Michael walked through three ways they've used the tool:

Use case 1: From unclear research brief to expedited launch

Researchers often receive chaotic, last-minute requests filled with conflicting stakeholder comments, vague goals, and tight deadlines. Charley Olman demonstrated how combining a custom Claude skill with Dscout MCP transforms a messy draft brief into a fully programmed, methodologically sound research projects in under ten minutes.

By embedding research best practices directly into the AI, the system cleans up conflicting requirements and pushes the study directly into Dscout.

An example of how to use this feature:

1. Upload the draft brief: Drop the messy or conflicting research brief directly into the Claude chat window.
2. Apply a specialized Claude skill: Prompt Claude to evaluate the brief using a pre-built research skill that incorporates your team's specific methodologies, quality standards, and guardrails.
3. Review flagged issues and generate document: Review the automatically flagged conflicts (such as redundant rating scales, participant cap mismatches, or device incompatibilities) and generate a clean, shareable study plan.
4. Build the study via Dscout MCP: Instruct Claude using Dscout MCP to take the refined study plan and automatically program a complete usability study—including screeners, prototype links, and varied question formats—directly inside Dscout.

Use case 2: Get live feedback directly on your designs

Instead of spending days manually organizing notes after interviews, researchers and designers can use an MCP to review participant reactions directly on prototype frames almost immediately after a session. 

Laura Oxenfeld shared how pairing Dscout MCP with a Figma MCP allows researchers to map live interview feedback directly onto visual design files, providing product teams with instant, visual insights.

An example of how to use this feature:

1. Take timestamped live notes: Log observations in Dscout’s live notes feature during the interview to help the AI understand which quotes and themes matter most.
2. Set up the Figma canvas: Copy the prototype screens into a dedicated frame in Figma where feedback will be placed.
3. Generate feedback callouts: Prompt Claude (using the Dscout MCP and Figma MCP) to read the transcript and notes, then place sticky notes and quotes directly onto the corresponding prototype elements grouped by sentiment (e.g., positive, negative, unclear, or new ideas).
4. Refine and train the AI: Review the generated frames, adjust any misplaced callouts, and feed those corrections back to Claude to continuously teach it your specific research standards.
5. Roll up across participants: After all sessions finish, instruct Claude to generate an aggregated cross-participant view in Figma that synthesizes overall feedback directly on the design screens for collaborative team review.

Use case 3: Tie insights to business initiatives

Connecting qualitative findings to broader company goals is a constant goal for research teams.

To help with this, Laura walked through how how running a UXR report (or raw interview transcripts) through an enterprise search tool (like Glean via its MCP) allows researchers to cross-reference user insights against an organization's entire internal knowledge stack—including Slack threads, Jira tickets, Confluence pages, and strategy documents.

An example of how to use this feature:

1. Analyze research findings: Input a finished UXR report or raw session transcripts into Claude, using the Dscout MCP to verify that no core themes were overlooked.
2. Query the enterprise stack: Prompt Claude using an enterprise search MCP (such as Glean) to scan internal tools for overlapping business initiatives, active roadmap items, or existing metrics related to your findings.
3. Categorize by business impact: Allow the AI to organize findings based on strategic relevance—such as items that block a live decision, shape an ongoing project, or offer directional strategy input.
4. Integrate context into deliverables: Add the surfaced business context, strategic links, and cross-functional team initiatives directly into your final report to show clear alignment with enterprise priorities.

Wrapping it up 

Adapting to new tools and workflows can sometimes feel overwhelming (so many new things every day!), but picking up new skills is a way to make sure your research remains fresh and relevant in this new era of AI. 

Interested in doing your own AI Show and Tell? We’d love to hear from you!

Curious about how to incorporate Dscout MCP into your workflow? Learn more here! Learn more here!

You may also like…

• Evaluating What Remains in Design as AI Breaks Boundaries
• When Rapid Prototyping with AI Causes Gaps, Try This Method
• Govern AI Quality at Scale with These Tips