For consultants asked about AI
AI Readiness Assessment From Your Notes
Your client's AI readiness, scored in six areas with a quote behind each score, plus use cases ranked by value and effort and a 90-day plan.
Free to start. No card needed.
How to do an AI readiness assessment yourself
Build the assessment on evidence, not a self-scored quiz. Interview the sponsor, the IT lead, whoever owns the key data, and two or three people who do the daily work. Ask for the systems list and any written AI policy. Then run a short anonymous staff survey, since AI use often stays hidden. Microsoft, which sells AI tools, and LinkedIn surveyed 31,000 knowledge workers in 2024. Of those using AI, 78% brought their own tools to work1, and 52% were reluctant to admit using it on their most important tasks.
Sort every fact under six headings: strategy and goals, data, technology, people and skills, governance and risk, and processes. Score each from 1 to 5 against a written definition of every level. The General Services Administration's AI Capability Maturity Model2 defines five, from “Initial / Ad Hoc” to “Optimized.” Score what exists today, not what is planned, and write the evidence beside each number. For governance, NIST's Govern function3 looks for policies, clear accountability and rules for third-party tools and data. Skip the overall average; name the weak area that blocks the most use cases.
Look for use cases in the client's own problems: frequent tasks whose output a person can check, such as reading forms or drafting replies. Keep judgment calls with people. In a Harvard Business School study4 of 758 BCG consultants, those using AI did work 40% higher in quality on tasks within AI's reach. On a judgment-heavy problem built on spreadsheets and interviews, they were 19 percentage points less likely to be right. Score each use case for value and effort, list its prerequisites, and pilot one or two quick wins against a measured baseline.
How it works
How River builds your AI readiness assessment
Start with what you already have. Paste your notes about the client, or drop up to 10 files, such as interview notes, a staff survey export, a systems list or a strategy document. Pick the client's size if you know it; if not, River infers it from headcount, sites or revenue in your notes and says which it used. Before writing, it reads every file, counts the interviews and survey responses, and copies each piece of evidence word for word under one of the six areas.
Each area gets a proposed score from 1 (not started) to 5 (leading), backed by a quote of 25 words or fewer and its source. Where your notes are silent, the area reads “Not enough evidence” instead of a guess. Plans never raise a score, and a score resting on one person or document is marked as one source. Size moves the bar. A one-page policy, a named owner and an approved tool list can earn a small firm a 4 for governance. An enterprise also needs an inventory of its AI systems.
Use cases come only from the client's own pain points and volumes, each scored for value and effort with its prerequisites. River never invents savings or prices. Risks are tied to a NIST function, a 90-day plan names an owner for every action, and the doc ends with the questions to ask next. More than 10 files? Condense the interview transcripts with interview synthesis first. To rank the AI risks alongside the client's other risks, use the business risk assessment template. When the client says yes, turn the plan into a consulting proposal.
How it works
From client notes to a readiness plan
Drop in your notes
Paste notes about the client, or drop interview notes and a survey export. Pick a size if you know it.
River weighs the evidence
It reads every file first, then scores each area only where a quote from your notes supports it.
Get your assessment
A scorecard, ranked use cases, risks, a 90-day plan and the questions to ask next, in one doc.
Share or go further
Edit and share the doc, or ask River for an interview guide, a client summary or a proposal.
| Area | Proposed score | Evidence |
|---|---|---|
| Strategy and goals | 2 of 5 | “Everyone wants AI, nobody owns it.” COO, interview 1 |
| Data | 2 of 5 | “Customer orders live in three spreadsheets.” IT lead, interview 3 |
| People and skills | Not enough evidence | Survey asked about tools, not skills or training |
| Governance and risk | 1 of 5 | No AI policy; 23 of 40 survey respondents use free chat tools |
First use case: typing up delivery notes · value 5, effort 2 · a quick win.
What you'll get
AI readiness scored from your notes, with use cases and a 90-day plan
- Six areas scored 1 to 5, each with the evidence behind it
- Up to eight AI use cases, ranked by value against effort
- Risks, a 90-day plan, the moves for months 4 to 12, and next questions
Before and after
Your client notes in. Scores you can defend.
Notes from five interviews at a regional freight company: the COO, the IT lead, two depot managers and the finance manager. Plus a 40-response staff survey in Excel and the client's one-page list of systems.
Readiness scorecard. Data: 2 of 5. “Customer orders live in three spreadsheets.” IT lead, interview 3. Governance and risk: 1 of 5. There is no AI policy, and 23 of 40 survey respondents use free chat tools at work (question 7). People and skills: not enough evidence. The survey asked which tools staff use, not about skills or training. Use case shortlist. 1. Typing delivery notes into the order system: value 5, effort 2, a quick win. Needs the scans in one shared inbox first. Days 1 to 30. Approve a one-page AI use policy and a short list of allowed tools. Owner: COO. Ask the client next. Who owns customer data today, and who can approve access to it?
Each score carries its quote and source, and an area your notes don't cover says so instead of guessing.
What lands in your doc
What's in your AI readiness assessment
- A readiness scorecard: six areas, each scored 1 to 5 with a word-for-word quote and its source
- “Not enough evidence” wherever your notes are silent, with the question that would settle it
- The binding constraint named instead of an average: the weak area that blocks the most use cases
- Use cases drawn only from the client's own problems, each scored for value and effort, with prerequisites
- Four to six risks, each tied to its evidence and to a NIST AI Risk Management Framework function
- A 90-day plan with owners, the moves for months 4 to 12, and the questions to ask next
Questions consultants ask
Common questions
What is an AI readiness assessment?
An AI readiness assessment is a structured review of whether an organization can use AI well and safely. It scores areas such as strategy, data, technology, skills, governance and processes, often on a 1 to 5 maturity scale, then names the gaps to close first. A useful one shows the evidence behind every score and ends with use cases and a plan, not just a number.
What questions should an AI readiness assessment ask?
Ask one or two factual questions per area. Who owns AI, and which business goal should it serve? Which systems hold the key data, and who can approve access? Which AI tools do staff use, and with what training? Is there a written AI policy and an approved tool list? Which tasks take the most time, and is that measured? River can write an interview guide for the gaps.
Which framework should an AI readiness assessment follow?
For governance and risk, NIST's AI Risk Management Framework3 is a sound base: it is voluntary and organized into four functions (Govern, Map, Measure and Manage). NIST published it in 2023 and is now revising it. For public-sector clients, add GAO's AI accountability framework5, which lists key practices for governance, data, performance and monitoring. River checks against NIST, but its assessment is not a certification or an audit.
How do I do an AI readiness assessment with River?
Paste your notes about the client into the box at the top of this page, or drop up to 10 files, such as interview notes or a survey export. Pick the client's size if you know it. River reads everything, scores each area from the evidence and writes the full assessment into your free River workspace, where you can edit it and share it.
Can I assess my own company's AI readiness?
Yes. Paste notes about your own organization instead of a client's: what you do, your systems and data, your team, and what you hope AI will do. The scoring works the same way. Areas you haven't described come back as questions, and those make a good agenda for a talk with your leadership team.
Is my client's material private?
Yes. Your notes and files stay in your own River workspace, private unless you share them. To write the assessment, River sends them to outside AI services, which work under data protection agreements and don't train their models on your content. River's privacy policy says it doesn't use your documents to train AI models either. Check your client agreement before you upload anything it restricts.
Is the AI readiness assessment free?
Yes, it is free to start, with no card needed. Your assessment opens in a River workspace, where you can edit it, share it or ask River to change it. Paid plans add more AI use and access to the most advanced models.
Your next step