royp.se

What It Is

Roy Persson is a personal, practice-oriented website focused on the intersection of AI, service design, and UX. The main article explains that when standard LLMs could not meet the author’s needs, they built a relatively simple AI-assisted analysis wiki for handling large volumes of interview material. Its core purpose is not to offer a standardized SaaS product, but to document how AI can be used in real service design and UX research work.

Core Capabilities and Model

The main capability mentioned on the site is RAG: interview content is split into searchable text chunks, indexed, and then used as source material for the model when questions are asked. The goal is not simply to get “faster answers,” but to make analysis more contextual and traceable. The article does not disclose the specific model, embedding approach, database, or tech stack, so its performance, stability, and scalability cannot be assessed.

Typical Use Cases and Output Quality

This approach is suitable for organizing interview data, analyzing UX research, extracting service design insights, and searching or summarizing large volumes of qualitative material in complex domains such as healthcare. Notably, the author emphasizes human in the loop: AI can identify patterns, but conclusions still need to be reviewed against the source material, context, and human judgment, while clearly distinguishing quotations, observations, and conclusions. This is a relatively robust quality-control approach for qualitative research. However, the site does not provide case outcomes, accuracy metrics, or user feedback, so output quality remains difficult to evaluate externally.

Pricing, Integrations, and Privacy

The article provides no commercial pricing, free quota, trial, payment method, API, or third-party integration information. It also does not disclose any data privacy or security mechanisms. Given that the application involves interview materials, privacy, permissions, and compliance should be critical considerations, but the current information is insufficient.

Pros, Cons, and Who It Is For

Its strengths are a focused use case and a practical methodology, especially its emphasis on source traceability and human review. Its weaknesses are unclear product maturity and a lack of pricing, deployment, technical, and security details. It is better suited for UX researchers, service designers, and qualitative researchers as a methodological reference rather than as a tool to purchase and deploy directly.

Access from China

There is no information about access from China, network availability, or payment support, so china_access can only be classified as unknown. If you need alternatives that can be put into production directly, consider Dovetail, NVivo, Atlas.ti, or using ChatGPT/Claude together with a self-built RAG workflow.