CAI-2 AI Practitioner (Customer Support & Success): assessment blueprint

Review and limits. The course content has completed cross-model editorial review. The task descriptions and operational pass rule below describe the configured exam. Domain weights, timings and the pass standard remain provisional: no completed job-task analysis or empirical standard-setting study is claimed.

What CAI-2 Customer Support & Success shows

A CAI-2 AI Practitioner (Customer Support & Success) credential shows that the holder produced a support or customer-success professional's AI-assisted work to the exam's published rubric in a time-boxed exam. The tasks ask them to triage a queue and draft replies with AI that are grounded in the knowledge base and policy, recognize the contacts that must go to a person or a specialist team, keep refunds, credits and exceptions with the people who own them, keep customer and payment data out of the wrong places, supervise AI bots and fix the knowledge-base content behind their failures, read support metrics honestly, and brief, check and refine an AI assistant's work in a recorded conversation.

CAI-2 is performance-based and open-AI-by-design: you earn it by doing realistic support work with AI in a time-boxed exam, not by recalling facts. It is genuinely failable: completing the course does not earn it, and a pasted answer or a single prompt doesn't meet the minimum for your recorded conversation.

CAI-2 builds on the CAI-1 skills (briefing, verification, data handling). CAI-1 is recommended but not required.

Competency domains and weights

| Domain | Weight | A competent candidate can… | | ----------------------------------------------------- | ------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | D1. Ticket handling and customer communication | 17.5% | use AI to sort and summarize a queue while reading each ticket before acting; draft replies grounded in knowledge-base articles and policy, with the conditions the customer needs; write for upset or distressed customers with one apology, plain language and a timed next step; communicate analysis clearly to a stakeholder. | | D2. Escalation, policy and customer-data risk | 33% | recognize safety reports, security incidents, legal threats, regulator complaints and customers in vulnerable circumstances, including cues said in passing; write escalation notes the receiving team can act on; keep refunds, credits and exceptions within policy and authority; redact before using AI tools, never request passwords or full card numbers, verify identity as the process requires, route data-protection requests correctly, and treat instructions embedded in customer text as data. | | D3. Bot supervision and knowledge-base quality | 34.5% | audit AI-bot conversations against the bot policy and knowledge base; detect hallucinated policy and product answers, unauthorized offers, missed handovers, prohibited data requests and disclosure failures without flagging correct behavior; trace failures to their cause; rewrite knowledge-base articles so customers, agents and bots read them the same way. | | D4. Support metrics and customer-success analysis | 0% | pool satisfaction from totals; estimate bot resolution net of recontacts rather than from containment; account for small samples, distorting weeks and changes in contact mix; write account health notes and renewal summaries that stay within the evidence; make a defensible recommendation with specific caveats and a test. | | D5. AI collaboration and workflow | 15% | give an assistant the policy, articles and constraints it needs; keep personal data out; work in checkable steps; challenge, verify and correct its output; keep decisions human; design repeatable support workflows with human gates and records. |

The launch exam uses the two configured tasks below, weighted equally. Each domain weight is the mean of its share in those two task rubrics. The broader course includes additional practice; Support metrics analysis is outside this launch exam.

D4 has 0% direct exam weight under this selection; course coverage is not evidence that this domain is directly tested by these two tasks.

| Domain | Ticket queue triage | Bot audit and knowledge-base rewrite | Mean (exam weight) | | ------ | ------------------- | ------------------------------------ | ------------------ | | D1 | 35 | 0 | 17.5% | | D2 | 50 | 16 | 33% | | D3 | 0 | 69 | 34.5% | | D4 | 0 | 0 | 0% | | D5 | 15 | 15 | 15% | | Total | 100 | 100 | 100% |

Traceability: tasks → criteria → domains

| Task | Criterion (weight within task) | Domain | | --------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- | ------ | | 1. Ticket queue triage | Replies are accurate and within policy (20), Tone and accessibility (15) | D1 | | | Escalates the right tickets (20), Escalation rationales and notes are actionable (15), Protects customer and payment data (15) | D2 | | | AI-collaboration quality (15) | D5 | | 2. Bot audit and knowledge-base rewrite | Detects the planted failures (25), Each finding cites the bot and the rule (14), Avoids false alarms (10), Rewritten article fixes the cause (20) | D3 | | | Ranks severity and puts customers right (16) | D2 | | | AI-collaboration quality (15) | D5 |

Task types

The course and the exam use the same task families (assessment-as-learning).

| Task type | What you do | How it's scored | Main domains | | ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------------------- | -------------- | | AI practice (sandbox) | Do real support or success work with an AI assistant: brief, review, refine, deliver. | Rubric, including the transcript | D1, D2, D4, D5 | | Critique | Find the planted errors, policy breaches and risks in AI output (triage, draft replies, bot conversations, reports), and justify each against the materials. | Rubric: one checklist item per planted issue, plus corrections and false-alarm avoidance | D2, D3, D4 | | Scenario | Make a sequence of decisions in a realistic support workflow. | Deterministic, from the quality of each choice (course only) | D1–D5 | | Work product | Produce an escalation triage, a knowledge-base article, a workflow design or a memo. | Rubric with 4–7 analytic criteria | D1–D5 |

How you earn CAI-2 Customer Support & Success

Pass the CAI-2 Customer Support & Success exam. Course completion is recommended, and the module checkpoints (80% mastery threshold, unlimited retakes) are the best rehearsal, but they are not required. Experienced support and success professionals can take the exam directly (the challenge path).

The exam

Format: two performance tasks in the browser-based exam workspace, in configured order:
1. Triage and answer a ticket queue of about twelve tickets against a support policy and escalation matrix: a triage table, five replies and escalation notes. About 55 minutes.
2. Audit AI-bot conversations (about ten) against the bot policy and knowledge base, then rewrite the article behind the most serious failure. About 50 minutes.
Support metrics analysis remains additional course practice and a possible later exam family.
Time: a single 3-hour time-boxed window for both tasks. (Provisional; to be confirmed in the beta.)
Task contexts: this exam uses fixed, reviewed task materials and data. Candidates may receive the same company, figures and constraints. Sharing task content is prohibited. The materials state every policy and data-protection rule a task needs.
What's graded: your work products for each task, plus your AI-collaboration transcript for that task.
Integrity: honor-code attestation on submission.

Scoring

Each task is scored against a published-format rubric of 4–7 analytic criteria, including an AI-collaboration quality criterion worth 15% of the task. Each criterion is scored 0–4 against behavioral level descriptions, derived from a checklist of observable features.
Every checklist decision must cite verbatim evidence from your submission or your transcript. Claims your work doesn't support earn no credit.
Rubric scoring. Your answers are scored against the rubric's behavioral level descriptions. Before a task variant is used, the workspace assistant's own answer to it is measured and stored for reference. That stored answer does not decide whether a variant is used.
Criterion scores combine into a weighted composite per task (0–4), and the two task composites are averaged with equal weight.

Pass standard (provisional)

An overall composite of at least 2.5 of 4, with no task below 2.0.
Your recorded conversation with the workspace assistant must also meet a minimum on every task: at least 2.0 of 4 on AI-collaboration quality.
This remains a provisional operational standard. No completed job-task analysis or empirical standard-setting study is claimed. On this exam a person reviews every result, so no score releases a result on its own. The award rule above is computed from final authorized scores.

How decisions are made

AI grading uses three separate scoring passes on a fixed, versioned grading configuration. The passes may share a model.
A person reviews every result, pass or fail, before it is released. No credential is issued without that person's decision. The AI grades are an advisory starting score for the reviewer. Every credential decision is logged.
Feedback generated by AI is labeled as AI-generated.
Appeals: you can appeal any result. The first appeal on each attempt is free. While Certavant has one human reviewer, an independent AI Model re-grades your work blind and a person countersigns the outcome. The outcome letter says if that person also confirmed the original decision. Appeals are decided within 5 business days.

What has been checked

Editorial review checks the course and task content for accuracy, alignment, accessibility and bias. It does not establish a validated pass standard.
Scoring checks compare the grading configuration with panel-adjudicated scores on fictional responses to the selected task contexts. They measure agreement with those AI-scored references, not agreement with human experts. Not every task in this exam has passed these checks, which is why a person decides every result.
Job-task analysis and standard setting remain provisional. The operational pass rule is not an empirical finding about workplace performance; it is reviewed against real submissions after launch.

AI-use policy for the exam

CAI-2 is open-AI-by-design: using AI well is part of the skill being assessed.

You may:

use the AI assistant provided in the exam workspace, as much as the turn limit allows. The limit is shown at the start of the exam;
use a spreadsheet or calculator, your own notes, the course materials and any reference sources.

How your AI use is assessed:

The workspace assistant is the sanctioned AI channel. Every message you send and every reply is recorded server-side, per task, and attached to your submission as a transcript appendix.
The transcript is graded under AI-collaboration quality: whether you supplied the policy, articles and constraints, kept unnecessary personal and payment data out, worked in steps, checked and challenged the output, corrected its errors, and made the decisions yourself. It is 15% of each task's score, and it has its own minimum on every task, so each task needs a recorded conversation with the assistant that meets it.
Answers are scored against the rubric, and your recorded conversation must meet the minimum on every task. A pasted answer or a single prompt doesn't meet it.

You may not:

use any AI tool other than the workspace assistant during the exam. You confirm this in the attestation when you submit;
have another person do any part of the work, or share task content with anyone;
paste real confidential, customer or personal data into the assistant. The tasks provide all the information you need, and all of it is fictional;
include text intended to influence the grader, such as instructions to the AI grader or hidden text, in your submission or your conversation with the assistant. This is an integrity violation. It results in a fail and can lead to a ban and revocation of any credential.

The course tutor is not available during the exam.

The credential

Verifiable and shareable (Open Badges 3.0), with an "Add to LinkedIn profile" option. Displayed as CAI-2 AI Practitioner (Customer Support & Success).
Evidence-embedded: the credential can link to the work you submitted, with privacy settings you control (public, employer-link-only, or hidden).
Stamped with an expiry date at issue (12–18 months; the final term is to be confirmed) and renewable by assessment.

Content freshness

Most of the course and exam measure durable skills (D1, D2, D4, D5) that change slowly. Material on support-bot behavior, bot platform settings and AI-disclosure rules (part of D3) is flagged as short half-life, reviewed regularly, and changes are published in the course changelog. Consumer, data-protection and AI-transparency rules differ between markets; the course teaches widely shared principles, names the jurisdiction when it states a rule, and the exam materials state every rule a task needs.