Rosenverse

Scaling User Research with AI: Continuous Discovery of User Needs in Minutes

Gold
Wednesday, September 10, 2025 • DesignOps Summit 2025
Share the love for this talk
Scaling User Research with AI: Continuous Discovery of User Needs in Minutes
Speakers: Joerg Beringer and Thomas Geis
Link:

Summary

Product Context Analyzer (PCA) is an AI-powered tool that automates LLM-based user research, enabling product teams to quickly discover and analyze the context in which a product is used. With just a single prompt or an uploaded interview transcript, PCA generates structured user research outputs that match the quality and format expected by professional design and product teams, enabling immediate integration into established workflows. Outputs include as-is scenarios, personas, empathy maps, user journey maps, desired outcomes, detailed task flows, user requirements, and user stories. By dramatically reducing the need for time- and labor-intensive primary research, PCA allows teams to continuously learn about user needs without requiring specialist skills or deep expertise in user research methods. This approach helps product teams move efficiently from early ideas to actionable design with a clear understanding of user requirements. In this session, we will demonstrate how PCA transforms a single input prompt into a structured knowledge graph that connects high-level strategic insights with detailed task flows and user stories.

Key Insights

  • AI can transform secondary user research by synthesizing vast existing information into actionable UX artifacts.

  • Jakob Nielsen recently endorsed a workflow prioritizing secondary research before primary user research.

  • Product Context Analyzer produces outputs like journey maps, task hierarchies, affinity diagrams, persona fact sheets, and user story maps automatically from a prompt or interview transcripts.

  • The tool supports input from either broad task descriptions or detailed interview transcripts.

  • The system structures user requirements in standardized, readable phrasing suitable for both UX and regulated industry contexts.

  • It incorporates concepts from Jobs-To-Be-Done and OOUX frameworks to offer desired outcomes and task objects.

  • Export options include CSV files and printer-friendly PDFs, facilitating easy sharing and further processing.

  • Currently, the system allows only one transcript upload per project and lacks multi-user shareable access but this is on the roadmap.

  • The AI insights link back to source data internally via knowledge graphs, ensuring traceability.

  • The tool is applicable across industries including medical devices, software, automotive, research institutes, and public services.

Notable Quotes

"The secondary first, primary second sequence may seem a numerical mismatch, but it is the way to go."

"Our product simply sucks everything out of the open AI system and processes it with a knowledge graph based on ISO standards."

"You enter a single prompt, like washing your own car, and the system analyzes the whole context and returns research artifacts."

"When designers describe their problem, they often describe a pain point, but you have to enter the user’s actual task."

"The output is not a requirements statement but a description of what is happening in the context."

"The task objects are central to a popular conceptual design approach called OOUX."

"Product managers appreciate the jobs to be done desired outcome statements because they are not overly detailed but actionable."

"This is real serious analysis, not pre-prepared or fake data."

"The system is live now analyzing audience-submitted tasks and producing results in minutes."

"Secondary research becomes a primary choice in user research in the age of AI."

Ask the Rosenbot
Emily DiLeo
Stronger Together: Lessons Learned from UX Research Ops
2024 • DesignOps Summit 2024
Gold
Victor Udoewa
Research in the Pluriverse
2023 • Advancing Research 2023
Gold
Christian Madsbjerg
Influencing Strategy
2020 • Advancing Research 2020
Gold
Catherine Dubut
Bridging Physical and Digital Spaces: Approaches to Retail Service Design
2021 • Enterprise Community
Ian Johnson
Latent Scope: Finding structure in unstructured data
2025 • Designing with AI 2025
Gold
Sha Hwang
The Lost Year
2021 • Design at Scale 2021
Gold
Maria Giudice
Becoming a Changemaker by Leading with Design
2023 • Advancing Research 2023
Gold
John Maeda
Making Sense of Enterprise UX
2016 • Enterprise UX 2016
Gold
Louis Rosenfeld
GenAI for UXers: A Rosenbot Demo and Discussion
2025 • DesignOps Summit 2025
Gold
Bria Alexander
Day 2 Welcome
2024 • DesignOps Summit 2024
Gold
Ross Smith
Breaking Barriers with Empathy
2017 • Enterprise Experience 2017
Gold
Ariel Kennan
Building a Design Culture
2017 • Enterprise Experience 2017
Gold
Ren Pope
Building Experiences for Knowledge Systems
2023 • Enterprise UX 2023
Gold
Mansi Gupta
Drawing from Feminist Practice to Make Inclusive Design Operational
2022 • DesignOps Summit 2022
Gold
Niko Laitinen
Adaptable Org Design for Resilient Times
2021 • Design at Scale 2021
Gold
Sheryl Cababa
Day 2 Panel
2024 • Designing with AI 2024
Gold

More Videos

Tom Armitage

"AI is like my coach. It’s like my bench player. It’s on my team, but you can’t rely on one player."

Tom Armitage Carla Diana Kanene Ayo Holder

Day 2 Panel: Looking ahead: Designing with AI in 2026

June 11, 2025

Victor Udoewa

"I want a story that moves and quakes me, palpitates my heart and shakes me."

Victor Udoewa

Research in the Pluriverse

March 29, 2023

Jon White

"We are not just doing research to find stuff out; we need to show a return on investment on everything."

Jon White Erin May

Unsticking Research for Better Information Flow

March 11, 2026

Matt Duignan

"Durable insights can emerge organically based on usage and referencing patterns in the system."

Matt Duignan

Atomizing Research: Trend or Trap

March 30, 2020

Shahrzad Samadzadeh

"Building relationships beyond roles, as people, makes everything else fall into place."

Shahrzad Samadzadeh

What Is My Value? Two Takes and Some Mistakes

January 8, 2024

Russ Unger

"Sharing a language of design gave us and our partners a consistent experience and the confidence to work together."

Russ Unger

Getting Out from Under Everyone: How to Escape the Paralysis of Getting Started

June 8, 2016

Elana Chapman

"Assistive technology is a tool; we need to make sure we enable that tool to do its job."

Elana Chapman Li Wen Huang Divyen Sanganee Annabel Weiner

Getting started with accessibility research

February 20, 2025

Jennifer Kong

"Users were initially excited but six months later expressed frustration and uncertainty about the tool’s usefulness."

Jennifer Kong

Journeying toward AI-assisted documentation in healthcare

June 5, 2024

Sheri Byrne-Haber

"You want to catch accessibility defects early through automated testing in your CI/CD pipeline, what we call shifting left."

Sheri Byrne-Haber

Accessibility at Scale

June 9, 2021