Tech-Driven Business

The episode surfaces a handful of real practitioner observations - requirements-gathering becoming the high-value skill as code-generation is commoditized, and point AI solutions creating context silos - but the conversation meanders with repetition, host restatements, and vague encouragement between each substantive point. Idea density per minute is low for an 18-minute runtime.

“It can really handle what would be major portions of work that we would sell as consultants in terms of hours and bags of hours, buckets of hours.”

“I'm seeing things be built like, ah, we wanted a generative AI app for this specific thing and we're going to get that deployed in two weeks or whatever. Actually, this is leading to some silos”

The framing of 'number of change orders' as a proxy for consulting quality is a mildly fresh take, and the honest breakdown of consulting moats is candid, but the core thesis - AI commoditises coding, judgment and requirements-gathering remain valuable - is a recycled argument circulating widely in 2024 tech commentary. No genuinely contrarian or first-principles arguments appear.

“how many change orders did you take to get to the finish line, right? I think that matters.”

“A lot of consulting companies, if you really think about the moats that they have, it's really nothing, right? It's either you're really large and you have a bench... And then number two, you have a bunch of slide where that speaks about methodologies, which are just derived from agile or whatever.”

Shashank is a genuine practitioner running a data and ML practice with real SAP delivery experience and acknowledged scar tissue from failed projects, which gives his commentary credibility. However, there is no indication of exceptional scale, seniority, or a distinctive track record beyond mid-level practice leadership at a small firm, limiting his caliber ceiling.

“Run a data and ML sort of practice focused around SAP”

“as somebody who's bid on so many projects, who's been part of so many RFPs, many of them not one”

Almost no hard data, named customers, real dollar figures with context, or timestamped examples are offered. The few numerical references ('eight sprints and maybe four,' 'couple hundred thousand dollar bill') are illustrative approximations with no sourcing, and SAP/Datasphere are name-dropped rather than analysed with evidence.

“a couple hundred thousand dollar bill, et cetera. That was a big part of the business.”

“you're able to wrap up a project that would have taken eight sprints and maybe four”

The host's questions are predictable and often leading or compound, and there is no instance of genuine pushback or a follow-up that presses for evidence. The host frequently restates the guest's point as affirmation rather than probing deeper, and several transitions are vague topic pivots rather than craft-driven follow-ups.

“Talk about the skills now. Consulting skills, there are some that are becoming more like commodity. There are still other skills which are highly sought after”

“That's a great way to look at it. It has to be context driven”