Diagnose

AI readiness assessment scored on how work actually happens.

An AI readiness assessment tells you whether AI spending will turn into results your finance team can verify. The Codos version is self-serve: 21+ questions across 8 areas in 5 to 8 minutes, with a scored report and a dollar estimate of the opportunity at the end. A 30-day diagnostic then turns the score into ranked opportunities with owners and a baseline your finance team signs.

Take the assessment
Self-serve assessment
5–8 minutes, 21+ questions, 8 areas, instant scored report
Built for
Companies of 300–5,000 people
Next step
30-day diagnostic: score to ranked opportunities and a signed baseline
Deployment
Your infrastructure: on-prem or your cloud
Backed bya16z Speedrun
Builders fromMcKinsey & CompanyMetaTesla
$5M+verified client impact

The measure

What an AI readiness assessment measures.

An AI readiness assessment measures whether specific workflows can move to AI in production: who owns the result, whether the data is reachable, whether the systems can be integrated, whether people will run the change, whether rules keep it safe, and whether finance agrees on cost and return. Data is one of six dimensions, not the whole test.

Readiness is not maturity. A maturity assessment grades how far along an adoption curve you are, and the grade can improve for years without a dollar of impact. A readiness assessment is forward-looking: it gates whether spending this quarter, on a specific workflow, will produce a result finance will sign.

Most AI spending fails this gate. MIT’s NANDA project found 95% of corporate GenAI pilots produce no P&L effect, and in a Gartner survey only 36% of CFOs were confident they could get meaningful value from AI. The difference is rarely the model. It is the six dimensions below, scored on evidence rather than intention.

Six dimensions of AI readiness, and what evidence counts
DimensionWhat it asksWhat evidence counts
Ownership and sponsorshipWho answers for AI resultsA named executive owner with budget for both software and process change, not a committee or an innovation team
Data and knowledgeWhether the inputs are reachableExportable data behind the top three workflows, and critical know-how written down rather than living in people’s heads
Systems and integrationWhether automations have somewhere to runAPIs that have actually been used, and a sanctioned runtime with credentials and logs IT stands behind
People and workflowWhether the change will be run, not resistedTeams asked where the friction is, and an agreement on how freed capacity gets used
Governance and riskWhether automations can run safelyWritten data rules, an evaluation step before go-live, and a switch back to the manual process that works in minutes
Finance and valueWhether results can be verifiedA baseline cost per workflow that finance agrees with, and impact verified outside the team that built the automation

Honest scoring

Why a questionnaire is not enough.

Almost every AI readiness assessment on the market is a multiple-choice questionnaire. The person filling it in is usually the sponsor, and sponsors overscore. “Our data is accessible” can mean a documented API, or one analyst who knows where the exports live.

The Codos assessment treats the questionnaire as a starting point, not a verdict. The self-serve score tells you where to look. The 30-day diagnostic then verifies it the way an audit would: AI-led interviews ask leaders and employees to walk through the last workflow that got stuck, and forward-deployed engineers validate data, controls, and integration effort on the real systems.

What happens after the score

From score to production.

The assessment is step one of the full AI transformation consulting engagement. Each step narrows the same question: which workflow moves first, and what is that worth.

01

Self-serve score.

21+ questions in 5 to 8 minutes across 8 areas: your company and organization, data and context, and six functions (engineering, sales, customer support, operations, marketing, product). You get a score per area, a dollar estimate of the opportunity, and the specific gaps behind it.

02

30-day diagnostic.

AI-led interviews with the executive sponsor, the owners of data, technology, security, and compliance, and the people doing the work. Engineers validate systems, data, and controls in parallel.

03

Workflow map and ranked backlog.

The diagnostic returns a map of how work actually happens, an ROI-ranked backlog of automation candidates, and a 30/60/90 plan with the financial mechanism behind each item. If you continue, the map becomes a company brain: a living ontology of people, workflows, and metrics.

04

Baseline with finance.

Before anything is built, the current cost of each first workflow is agreed with your finance team. Owner first, baseline second, tools last.

05

First workflow runs.

Forward-deployed engineers build, evaluate, and ship the top automation, and the AI Office reviews impact, owners, and blockers with the C-suite every week.

When to run one: before the first pilot, before buying tools, and before any automation touches production.

Try it on paper

Five questions, and who this is for.

A fast version you can run in your head, one question from five of the six dimensions:

  • Can your CEO name the three workflows AI should change first?
  • Can you export the data behind your top three workflows without filing a vendor ticket?
  • Is there a sanctioned place for automations to run, with infrastructure, credentials, and logs IT stands behind?
  • Is there an agreement on how freed capacity gets used: redeployment, a backfill freeze, or reduction?
  • Is there a baseline cost for each candidate workflow, and does finance agree with it?

Silence on any of them is a real gap.

The assessment is calibrated for companies of 300–5,000 people: PE-backed operators, finance and technology businesses, and other organizations where the cost base is real and pilots keep stalling. Deployment is on your infrastructure, so nothing the diagnostic touches leaves your control.

Questions buyers ask

Before you take it.

By scoring evidence, not sentiment, across six dimensions: ownership, data, systems, people, governance, and value. The self-serve assessment produces the score in 5 to 8 minutes. The 30-day diagnostic then verifies it through AI-led interviews and engineer validation of real systems and data.

A readiness assessment is forward-looking: can we deploy now, and what is in the way? A maturity assessment grades how far along an adoption curve you are. Maturity levels can rise for years without a result finance will sign.

The self-serve version takes 5 to 8 minutes. The Codos diagnostic (AI-led interviews with the workforce, in-person interviews with executives, and a check of the real systems) takes 30 days. Pick the format by the decision you need to make.

Turn the score into a sequence: fix the gaps that block the first workflow, rank candidate workflows by ROI, and take one to production with a named owner and a baseline your finance team has agreed.

Start with the score

Score your readiness in eight minutes.

The self-serve assessment is free and specific enough to argue with. If the score confirms what you suspected, the 30-day diagnostic turns it into a workflow map, ranked opportunities, and a baseline your finance team signs.

Take the assessment
founders@codos.ai