Code Story: Insights from Startup Tech Leaders

The episode contains a handful of genuinely useful operational details - 8 concurrent real-time AI models, a compliance engine, a latency-driven UX failure with ElevenLabs - but these are spread thin across heavy origin-story narration, generic hiring platitudes, and motivational closing advice that add little for operators.

“there's around eight, eight different AI models that are all running in real time”

“if you use 11 labs off the shelf every now and then in let's say one in 20 interviews, the AI would scream”

The observation that resumes are now a dead signal and interviews are 'the last signal left' is a crisp, grounded contrarian claim; however, most of the episode defaults to recycled startup playbook content - ship faster, play the long game, hire senior hands-on people - without offering a novel angle on any of it.

“most recruitment teams weren't looking at resumes anymore because they decided there's not a lot of value in that because they're all just fake now”

“the key advice that I would tell any entrepreneur is this is a long game”

Arsham is a genuine practitioner - former Head of ML at a health-tech company, Amazon background, and currently running a 10-person team processing one million interviews per year for enterprise clients including a top-tier automotive OEM; he's not a thought leader, but the company is still early-stage so scale and proof points are limited.

“I was the head of machine learning at a prior company”

“we're doing around a million interviews a year now, so still very lean”

The episode is reasonably specific by podcast standards - citing team size (10), interview volume (1M/year, tested to 10M), AI model count (8), development timeline (9 months on voice), and 1-in-20 screaming failure rate - though several future claims (100M interviews next year, 48-hour hiring cycle in 2 years) are aspirational and uncorroborated.

“we're doing around a million interviews a year now”

“if we're doing a million interviews a year right now, we've actually tested all the way up to 10 million”

The host relies almost entirely on pre-scripted transitions ('let's dive into,' 'let's flip the script,' 'let's move forward') and never pushes back on any claim - the 100M-interviews-a-year projection, the 48-hour hiring cycle, or the bold market-size assumptions all pass unchallenged, making this feel more like a produced PR feature than a genuine interview.

“Let's dive into what you would consider the MVP for ribbon then”

“So as you step out on the balcony and look across all that you've built thus far with ribbon, what are you most proud of?”