Season 2 is here
We are back with a second season of Skimlecast — more episodes, new topics, and the same two people who cannot stop talking about qualitative data and what AI can and cannot do with it. Season 2 opens with three episodes: the origin of Skimle (properly told for the first time), how AI is changing academic research practice, and a wider conversation about where knowledge work is going.
Episode 1: Welcome to Season 2 — where Skimle came from and where it is heading
We have been asked plenty of times how Skimle actually came to exist. This episode is the full story. It starts with a failed experiment: Henri was working with a Finnish government ministry on a retrieval-augmented generation (RAG) system to make sense of large document corpora. The technology worked in demos and fell apart in practice. The documents were too unstructured, the query interface too blunt, and the outputs impossible to verify. What the researchers actually needed was not a search engine over a pile of documents — they needed a way to read, code, and synthesise qualitative material at the speed that the volume of data demanded.
That failure became the starting point for Skimle. Henri built a prototype focused on the analysis workflow rather than the retrieval problem. Olli joined shortly after, and the two of them have spent the time since trying to answer a deceptively simple question: what would Excel look like if it had been built for knowledge workers who work with text instead of numbers?
We talk about the product vision — horizontal rather than vertical, broad across research contexts rather than optimised for one domain — and walk through every major feature in the current platform. That includes AI-powered transcription, Skimle Ask for running structured AI interviews at scale, Skimle Anonymise for GDPR-compliant de-identification, the automatic thematic analysis engine, metadata variables for segmented analysis, agentic AI chat that can read and edit your project data, and REFI-QDA export for researchers who want to move results into NVivo or ATLAS.ti. We also announce a focused usability month — a period of targeted improvements driven entirely by the friction points we are hearing from users.
If you are new to Skimle and want a reference point for what the platform does, how Skimle works is the companion read, and the academic researchers and consultants and investors use case pages show how different communities are putting it to work.
Episode 2: Using AI in academic research — dirty secrets, qualitative surveys, and what actually works
This episode is for academic researchers who are quietly using AI in their work and not quite sure what they are supposed to admit in public. The dirty secret Henri refers to in the title is that many qualitative researchers are already using large language models as part of their workflow — for transcription, for initial coding passes, for organising fieldnotes — while the methodological literature has barely started to catch up. The norms around what counts as rigorous, what needs to be disclosed, and what is simply off-limits are still being worked out.
We go through where AI genuinely helps in qualitative research (the volume problem, the consistency problem, the speed problem) and where it does not (interpretive judgement, theoretical development, the kind of contextual reading that requires the researcher to bring meaning to the data rather than extract it). The hallucinations, context windows, and black-box outputs post covers the failure modes in detail if you want the full argument.
The second half of the episode deals with AI-assisted interviews, also known as qualitative surveys and explain our approach with Skimle Ask.
Episode 3: The future of knowledge work — AI slop or greater sophistication?
The provocation for this episode is a pattern we are both noticing: as AI-generated content becomes easier to produce, there is a real risk that the outputs of knowledge work — reports, analyses, research findings, strategic recommendations — converge toward a kind of fluent mediocrity. Everything sounds well-written. Nothing says anything. We call this AI slop: content that passes surface-level quality checks while being empty underneath.
The counter-argument, which we spend most of the episode developing, is that AI tools could just as plausibly drive the opposite: a shift toward greater analytical sophistication, where the time previously consumed by mechanical tasks (transcription, coding, formatting, summarising) is freed up for the parts of knowledge work that actually require a human — interpretation, judgement, synthesis across domains, and the ability to ask the question that was not on the original list. The outcome depends almost entirely on how the tools are designed and how practitioners choose to use them.
This connects to a thread we have developed across several posts. The death of SaaS or a renaissance of better software piece looks at how AI is restructuring the software layer underneath knowledge work. Two-way transparency is our argument for why traceability — the ability to follow every insight back to its source — is what separates analysis from noise. And why RAG approaches fall short for structured qualitative analysis gets into the technical reasons why the architecture matters, not just the interface.
The question we finish on — whether AI will make knowledge workers more sophisticated or less — does not have a clean answer. But it is the right question to be asking, and we think the people in this field are better positioned than most to shape which direction it goes.
About Skimle and Skimlecast
Read more about Skimlecast and watch Episode 1 here! You can watch all our episodes on our Spotify or YouTube channels. If you have any comments, feedback, topics you would want us to cover or other things you want to share, please connect with us through our form or via email through olli@skimle.com.
If you are interested in using Skimle, check out how Skimle works. You can also try Skimle for free and see how AI-assisted qualitative analysis handles everything from academic interview data to customer feedback at scale.
Meet the cast
Henri Schildt is a Professor of Strategy at Aalto University School of Business and co-founder of Skimle. He has published more than a dozen peer-reviewed articles using qualitative methods, including work in Academy of Management Journal, Organisation Science, and Strategic Management Journal. His research focuses on organisational strategy, innovation, and qualitative methodology. Google Scholar profile
Olli Salo is a former Partner at McKinsey & Company where he spent 18 years helping clients understand the markets and themselves, develop winning strategies and improve their operating models. He has done over 1000 client interviews and published over 10 articles on McKinsey.com and beyond. LinkedIn profile