First Commit

The episode surfaces a few genuinely useful technical observations - the non-determinism gap in Temporal/Conductor, the glass-box approach to MCP tool invocation, and separation of execution from plan generation - but large stretches are consumed by career biography, co-founder relationship advice, and generic hiring philosophy that add little actionable value for a B2B operator.

“around 2023, we knew that companies like Temporal and Conductor have solved for a lot of the durability and core orchestration problems. But at the time, we noticed that they haven't solved for the non-determinism that comes with AI models.”

“we take a very glass box approach where you know exactly which tools could be invoked or will be invoked, and you can select and remove like certain tools versus kind of like pulling in an entire server and giving it full control with natural language”

The ToolFormer-to-function-calling intellectual lineage and the prediction of a near-term in-house AI build correction/collapse are modestly contrarian and grounded, but the bulk of the episode recycles widely-circulated startup wisdom about co-founder fit, MVP redefinition for enterprise, and 'hire for fundamentals.'

“that paper was almost like the inception of a lot of what we see right now: things like function calling, tool calling, MCP and agent”

“there's a little bit of an overcorrection of a lot of folks building in-house, and then there's gonna be like market correction. It's like almost like the reinvention of SAS”

Mayada Gonimah is a genuine technical practitioner with credible depth - infrastructure engineering at Goldman Sachs, PCI-compliant payments at the New York Times, and MLOps/model evaluation at Palantir serving the DoD - making her clearly operator-class rather than a thought-leader, though Thread AI is early-stage and the interview surfaces little proof of scale.

“when I started my career at Goldman Sachs, I was primarily an infrastructure backend engineer, building software for various business divisions and prime brokerage and futures trading”

“the work that we've done at Palantir has been working with not just large government agencies like the defense, but also working with other large enterprises”

The episode names real technologies (Temporal, Conductor, MCP, ToolFormer paper from Meta, all three hyperscaler clouds, GDPR, HIPAA) and concrete use cases (invoice processing, insurance claim triage, diligence reporting) with a candid data point on sales cycle variance, but Thread AI's own metrics - ARR, customer count, agent performance benchmarks - are entirely absent.

“we were also really inspired by that tool former paper that came out of Meta, I think early 2023”

“We've had cases where some contracts were closed within a week, other cases where it's taken us over two years to redline a particular contract”

The hosts ask a few relevant operational questions (customer qualification, services vs. software tradeoff, what's gone wrong in deployment) but consistently drop promising threads - when the guest predicts Q3/Q4 security collapses the host pivots immediately to hiring - and several questions lean toward founder-narrative psychology rather than technically probing the product or market.

“Have you seen anything gone wrong yet when you've been deploying using these autonomous systems, or is it too early?”

“How have you learned what customers to say no to in the sense of you're working with pretty risk-averse buyers in financial services, public safety, healthcare?”