DeepRec.AI

There are occasional useful operational specifics - quantified loss ranges, a concrete ICP definition, and an interesting AI-agent-orchestration vision - but the majority of the episode is padded with product walk-throughs, origin storytelling, and generic SME pain-point framing that offers little a B2B operator couldn't infer themselves.

“90 to 95% of the time, that is because an asset, a physical piece of the railway has gone offline”

“they might lose anywhere from, you know, £10,000 for a missed delivery, all the way up to £100,000”

The 'AI agents as logistics orchestration layer' framing is topical but mirrors widespread 2024 AI discourse, and the Amazon-for-enterprise-logistics analogy is a well-worn comparison. No contrarian or first-principles arguments appear; the episode recycles standard 'legacy tools vs. AI' positioning without interrogating it.

“Sort of the way that Amazon does it for us as a retail customer is, we imagine the same thing, logistics”

“a lot of the systems in logistics are disconnected”

Selby is a genuine practitioner with 8+ years in logistics and manufacturing and a verifiable Network Rail deployment, which gives credibility above a pure thought-leader. However, Mapify is only six months old at recording, limiting demonstrated scale and making many claims forward-looking speculation rather than hard-won experience.

“Eight plus years working in logistics and manufacturing”

“we founded in April of this year”

A handful of concrete figures anchor the episode - ICP revenue bands, per-incident loss ranges, and a named mapping vendor - but the bulk of the content is qualitative product description without supporting data, customer case studies, or independently verifiable metrics.

“they're generally generating anywhere from a million pounds in turnover to 10, 20 million pounds a year”

“we have users that are doing anywhere from 200,000 to a million pounds month in revenue”

The host repeatedly validates rather than interrogates ('Amazing. Amazing.' appears multiple times, alongside 'really, really excellent'), asks multi-part leading questions, and never challenges a single claim about product efficacy, competitive differentiation, or business traction. The conversation functions as a promotional interview rather than a substantive probe.

“Amazing. Amazing. I'm really excited about this particular episode”

“I think it's, I think it's excellent”