The IoT Podcast

The episode contains a handful of concrete observations (UWB talent scarcity, Poland's embedded engineer pool, the 'no bugs, no maintenance cost' value proposition) but is padded heavily with high-level positioning talk and business clichés. The ratio of genuinely non-obvious claims to filler is low for a 49-minute runtime.

“once you guys did the code there were no bugs to fix so there was no maintenance cost associated to this and no iterations needed. It was a certified ready to go code for a very niche technology”

“replacing a team or having two teams from uh, Western Europe or US versus a mix of 90% Polish engineers with some addition of guys, um, from Asia. Basically the output was 97% um, of the same quality”

Almost all frameworks deployed here are standard B2B sales orthodoxy - ICP analysis, value-based selling, thought leadership, champion identification. The lone semi-original observation (embedded engineers learn AI more readily than vice versa) is sensible but not novel, and the humanoid RTLS application is speculative rather than argued from first principles.

“It's easier to learn I think, AI when you're an embedded engineer than other way around. So you need to look from the people who already know embedded kind of are learning AI”

“the key is we start with the client problem. Uh, we take very vague uh, requirements and it's a mind shift. So you're not starting with a technical solution, you're starting with the business problem”

Dominik has relevant enterprise-side experience (Ecolab RTLS pricing, Swiss OEM/hearing-aids commercial roles) and brings legitimate commercial perspective to an engineering-led company, but he joined NeedCode only two months before recording and the conversation is largely promotional positioning rather than battle-tested operational depth.

“my last episode in the corporate world was at Ecolab where I was doing real time location system pricing and commercial strategy”

“I joined NITCO two months ago as a full time um engagement leaving uh my corporate world”

There are named companies (Qorvo, Embedded World, Ecolab), rough figures (12,000+ Polish embedded engineers, 150-200 partner conversations, 55-meeting enterprise sales cycles, 97% quality benchmark), and two named case studies (boat elevator, corn-field monitoring), but the Stanford research is uncited, most numbers are approximate, and case study detail is deliberately withheld.

“he was looking for ultra wideband capabilities for more than two years in the U.S. he said there is um, kind of one place that didn't disclose where you can find those capabilities”

“I think there are more than 12,000 engineers in Poland that are related to embedded”

The host structures the conversation reasonably and covers relevant ground (AI threat, certification, ICP, cultural dynamics), but every answer is validated rather than probed - there is no pushback, no challenge to vague claims like the uncited Stanford study, and questions frequently start with affirmative summaries that let the guest off the hook.

“That's great. And to give us maybe a uh, bit of context around that”

“Yeah, absolutely. And I think in the context uh, of what you were saying before of going from it just being a conversation around cost”