Every board asks the CTO the same three questions every quarter.

Which initiatives shipped? What did each one cost? Which one paid off?

Most CTOs cannot answer them without scrambling. You spend a Friday afternoon stitching the answer together from Jira, payroll, and finance, then walk into Monday’s board meeting hoping the numbers add up. Usually they don’t, and the board can tell.

The 51-second short film above says it best. The rest of this post is what we have been shipping at InteliG over the past two weeks to make sure you don’t have to scramble like that anymore.

Why most CTOs can’t answer the three questions

The board’s three questions sound like business basics. Each one quietly requires you to chain three different systems that do not speak to each other.

Which initiatives shipped?

Jira has labels. They are inconsistent and often empty. Git has the commits. They are not tagged to the initiative. So you ask your team leads to remember what was tied to which OKR and hope the right thing surfaces in the standup before Monday.

What did each one cost?

Payroll knows what engineers cost. Time tracking would tell you where they spent their hours, except nobody does time tracking honestly. Finance knows the budget at the team level, not the initiative level. So the answer rounds to “all of engineering cost what all of engineering costs,” which is not an answer.

Which one paid off?

Revenue lives in a CRM. Product analytics live somewhere else. The initiative that drove the win did not get tagged in either. So the connection between engineering work and business outcome is a story you tell, not a number you show.

Three questions. Three systems that do not connect. No clean answer. Every CTO knows the feeling.

The last two weeks at InteliG

We have been shipping pieces that turn each of those questions from a story into a number. Here is what landed.

Autonomous Action shipped to main

Cognis stopped being just an advisor and started being an operator. Reversible writes auto-run inline as Cognis reasons through a problem. Destructive writes still gate behind a one-click confirm card. The gating axis is reversibility plus boundary, not a question of trust.

Practically: when Cognis identifies that a commit needs to be linked to an initiative, the link gets created. When Cognis spots that a meeting belongs to a strategic theme, the connection gets made. You are not approving a queue of suggestions every morning. The chain builds itself, and you see the audit trail when you need it.

The Cognis Brain shipped

A 3D anatomical model that lights up by pillar as Cognis reasons. The intelligence layer is finally visible. Strategy lights up when initiatives are being parsed. Code lights up when commits are being chained. Finance lights up when cost is being attributed. Action lights up when something autonomous is running.

The brain is not a gimmick. It is an honest visualization of what the AI is actually doing, which is something every other platform leaves opaque on purpose. You can see when Cognis is overconfident in one pillar and quiet in another, and you can use that signal to ask the right next question.

CTO Briefing went default-on

Every exec at every InteliG org now gets a weekly board-ready brief, automatically, with full provenance. No setting up, no opt-in, no maintenance metadata. The brief covers what shipped, what cost what, what is blocked, and what the team should highlight to the board.

The brief has receipts. Every claim links back to the commit SHA, the meeting, the PR, the finance line, or the initiative record it came from. The brief does not have to convince the board. The receipts do.

The AI Maturity Framework went public

We published the AI Maturity Framework on this site this week. Five pillars, scored zero to four, licensed CC-BY. It is the same rubric we use when we audit an engineering organization.

The pillars are augmented engineers, agentic loops, standards as code, memory and context, and MCP and tool use. Score yourself on all five. Most teams we see land at two-two-one-one-two and do not realize they are bleeding leverage on the invisible pillars.

The framework is free. The methodology is open. The platform is the product.

The AI Portfolio Audit launched

A consulting offering for PE firms that runs the rubric across their portfolios. We deliver two artifacts per engagement: a strategic brief for the deal team (per-engineer productivity rankings, named insourcing recommendations, asset ownership, transition de-risking plan) and an onboarding briefing for the new CTO at the portfolio company (operational survival kit, anonymized at the individual level so the new CTO retains authority over personnel decisions).

First Phase 0 with a PE firm is free. After that, $3,500 per portfolio company. We launched it because the framework alone gets you a self-score. The audit gets you the truth.

How the stack answers the three questions

Now look at the three questions again, with the stack in mind.

Which initiatives shipped? Atomic Code Linking chains every commit to the initiative it served. Initiative Discovery reads the commits and tells you what your initiatives actually are, derived from the work rather than declared in an OKR doc. The Cognis Brain visualizes which initiatives are alive in the codebase right now. The CTO Briefing has the list ready for Monday morning.

What did each one cost? Commit-level time scoring, multiplied by engineer fully-loaded cost, attributed to the initiative each commit belongs to. Plus the meetings about that initiative. Plus the AI and token spend on agents working on it. Cognis Memory remembers the chain across sprints. You do not see “what engineering cost.” You see “what each initiative cost.” Per-initiative is the only resolution that matters to the board.

Which one paid off? Once cost lives at the initiative level, ROI follows. You compare cost per initiative against outcome per initiative (revenue contribution, retention, customer signal). The initiative that paid off is the one with the highest outcome over cost. Not the one the team likes telling stories about. The one the data says.

That is why the film opens on three questions and ends on verdicts, not vibes. The verdicts are the answers. The vibes are what most CTOs walk in with.

Why “shows” beats “pitches”

In the film, two CTOs walk into the same boardroom on the same Monday. One pitches. The other shows.

The CTO who pitches is the CTO with the verbs. We’re focused on. We’re prioritizing. We’re investing in. We’re driving toward. All present tense. All aspirational. No past tense. No numbers.

The CTO who shows is the CTO with the nouns. Initiative A shipped in eight weeks. Cost $230k in engineering time. Drove $1.4M in net new ARR. Past tense. Specific. Receipts attached.

The board does not trust verbs. The board trusts nouns. The CTO who shows up with nouns owns the room. The CTO who shows up with verbs ends the meeting fielding harder questions for next quarter and wondering why the deal lead keeps “checking in” between meetings.

That is the whole shift. That is what we have been shipping toward.

What to do next

Watch the film above. It is 51 seconds. It says it better than this post can.

If you want the framework we use to assess engineering teams, it is here. Free, CC-BY, score yourself on the five pillars in an afternoon.

If you run a PE firm and want the AI Portfolio Audit, the first one is free for the first PE firm in each region. Send me a note: levi@intelig.ai.

If you want InteliG itself running against your engineering org, it is at intelig.ai.

The board asks. You don’t pitch. You show.

That is what we have been building.

The Three Questions Every Board Asks the CTO (And the Stack That Finally Answers)