Kirby Montgomery says the first question about a good-looking engineering dashboard is who actually uses it and what decision it changes. He and Jim Milbury describe dashboards or cloud artifacts vibe-coded the night before diligence, hosted locally, and never shared with the C-suite or board.
Dan Bender says tech diligence used to be a confirmatory exercise three weeks before signing. It is now split and shifted left: an AI-driven first phase asks whether the company can defend against AI disruption and use it offensively, before the team goes deeper into architecture and cyber.
Kirby Montgomery says the audit reviews AI engineering and the product layer because many companies built useless chatbots or AI features nobody uses. For fine-tuned models, MCP servers, or other AI features, he asks who uses it, what outcome it drives, and what it costs per unit.
Dan Bender cites a statistic that less than 30% of tokens are consumed by frontier models; most requests can go to open-weight or smaller models. He says a key capability, probed in diligence and monitoring, is whether teams consistently choose when a frontier model is worth it.
Jim Milbury applies Little's Law: starting more work does not mean finishing more work. With AI he can create far more projects, but his finished projects have not increased; he calls that a discipline and prioritization problem, not a tooling win, and says throwing away dead-end prototypes is acceptable.
Jim Milbury argues you cannot rank individual engineers without breaking the measurement, especially when a board sees a low-ranked engineer who spent the quarter on architecture or customer work. He and Dan Bender agree that once a number decides pay, it stops telling the truth and human judgment must remain in the loop.
Dan Bender says he often finds the CTO has no idea what the private equity value creation plan is when a deal closes. Commercial strategy and engineering realities sit apart, and agentic coding makes that gap more dangerous because feature costs fall while token spend and total ownership stay unclear.
Dan Bender says the best code often comes from deleting code through refactoring, so lines of code is a poor proxy for progress. He argues token cost is a more real commercial constraint now, because companies cannot burn through tokens without a strategy or good engineering harness.
Kirby Montgomery says AI lets teams build software faster, so the old constraint of hiring more engineers fades. The mental model shifts to whether the system is stable and can scale with the plan, such as doubling the company or cutting time for a new customer to go live.
Kirby Montgomery describes a CTO whose CEO read about a 'dark factory' where robots claim code is tested. An investor holding the company for five years hears a reliability black box: if it breaks, customers cannot be served, and Dan Bender notes AI-generated code's total cost is still unknown.
Kirby Montgomery describes a deal where the company showed a beautiful next-generation platform built entirely with vibe coding and no reality. He told investors to pay for today's software, not the vision, and Jim Milbury adds that the future gap is usually scalability and reliability, not features.
Kirby Montgomery calls DORA only partial: speed and stability matter, but the best teams pair them with outcomes like feature usage, cost to serve, or revenue. Dan Bender adds that DORA fits modern startups better than private equity firms still running Delphi or fat-client apps they cannot deploy often.
Kirby Montgomery says monthly or quarterly monitoring must pair captured data with context, so non-technical investors understand why numbers moved. Code & Co. tells engineering teams it is not there for gotchas, but to translate their story; if the work is helpful, engineers lean in rather than game the metrics.
Kirby Montgomery says only one project this year had no material change to the product and technology value creation plan. Yet quarterly business reviews often do not know the product roadmap or whether the post-close technology honey-do list was done, turning the plan into a spreadsheet that nobody tracks.
