--- title: Connect Evaluation Checks (Oct 2026): Warnings Don't Stop Activation description: On October 9, 2026 Amazon Connect Customer added Save and validate in 8 Regions. Structural errors block activation. Gen AI warnings do not. The Gen AI check is capped at 3 in parallel and 30 per hour. url: https://www.factualminds.com/blog/amazon-connect-evaluation-form-automated-checks-2026/ datePublished: 2026-10-10T00:00:00.000Z dateModified: 2026-10-10T00:00:00.000Z author: palaniappan-p category: Cloud Architecture tags: amazon-connect, contact-center, evaluation-forms, quality-management, aws --- # Connect Evaluation Checks (Oct 2026): Warnings Don't Stop Activation > On October 9, 2026 Amazon Connect Customer added Save and validate in 8 Regions. Structural errors block activation. Gen AI warnings do not. The Gen AI check is capped at 3 in parallel and 30 per hour. On **October 9, 2026**, AWS [added automated checks that suggest edits to Amazon Connect Customer performance evaluation forms](https://aws.amazon.com/about-aws/whats-new/2026/10/amazon-connect-customer-automated-checks-evaluation-forms/). In the admin website, **Save and validate** runs two stages. Structural problems return as **Errors** and block activation. If the form has generative AI questions and no structural errors, those questions are checked against **11** best practices and return as **Warnings**. Warnings do not block activation. The generative AI stage allows **3** validations in parallel and **30** per hour on each instance. The check is in **8** Regions. AWS's own example is the question "Did the agent provide a sales disclosure?" The check asks the manager to name the exact disclosure. Activate with that warning still open, and the model can mark the question Yes because the transcript contains the word "disclosure." Cases CSV export, shipped August 4, 2026 in **10** Regions, is a different control. See [Connect Cases CSV export](/blog/amazon-connect-cases-csv-export-agent-workspace-2026/). Broader design (flows, workspace, Cases) stays on [Amazon Connect consulting](/services/amazon-connect/). > **Modeled comparison (not a cited client)** — One sales form, two gates. Gate A is structural: section and question limits, weights, scoring consistency, unique IDs. An error stops **Activate**. Gate B is the **11** generative AI checks. A warning does not stop **Activate**. The numbers here are AWS's **8** Regions, **11** checks, and the **3** / **30** rate limit from the [evaluation form admin guide, Step 9](https://docs.aws.amazon.com/connect/latest/adminguide/create-evaluation-forms.html). We have not published a before-and-after accuracy figure for a live queue. --- ## What shipped on October 9, 2026 | Piece | What AWS documents | | --- | --- | | Entry | Evaluation form → **Save** → **Save and validate** | | Stage 1 | Structural and configuration rules. Findings are **Errors**. They block activation. | | Stage 2 | Runs only when stage 1 is clean and the form has Gen AI questions. Findings are **Warnings**. | | Rate limit | **3** Gen AI validations in parallel, **30** per hour, per instance. Structural checks are outside that cap. | | Resolved | Hides a finding in the side panel. Does not re-run validation and does not change activation state. | | Regions | **8**, listed in the FAQ. Seoul and Cape Town are absent. | | Permission | Analytics and Optimization → Evaluation forms → manage form definitions → **Create** | A green banner means the pass found no recommendations. Findings reopen from the **Findings** control beside a question, or from **Save and validate** again. ![Authoring flow for Amazon Connect evaluation forms: a QA manager opens the form, Save and validate splits structural errors from Gen AI warnings, and activation leads to the analytics rule that submits evaluations.](../../assets/images/blog/amazon-connect-evaluation-form-automated-checks-2026-01.webp) _Figure: Save and validate, then Activate. [Open the draw.io](https://www.factualminds.com/examples/architecture-blog-2026/connect-evaluation-form-validation/connect-evaluation-form-authoring.drawio)._ --- ## The 11 checks on a Gen AI question AWS lists these for every question automated with generative AI. Warnings from this list do not block activation. Treat them as the quality bar anyway. | Check | What fails it | | --- | --- | | Question phrasing | Title is a heading or a statement, not a full question ending in a question mark | | Instructions present | No instructions telling the model how to answer | | Answer option language | Acronyms or abbreviations in the options | | Answer option conciseness | Options carry commentary or conditions instead of a short label | | Transcript answerability | The answer needs data outside the transcript and the instructions | | Positive action framing | The question asks the model to detect that something did not happen | | Plain language | Instructions use abbreviations, acronyms, or in-house jargon | | Spelling | Misspellings in the title, instructions, or options | | External systems | Instructions mention a screen or state the transcript cannot show | | Non-textual cues | Volume, pitch, speaking speed, or vocal tone. Professionalism and empathy are allowed | | PII values | A specific personal-data value appears in the question or instructions. Scoring whether the agent handled PII correctly is allowed | --- ## Clear Warnings before Activate **Clear every Warning, then Activate.** AWS still lets you activate with Warnings open. Another validation pass shares the **3**-parallel and **30**-per-hour cap, so a bulk cleanup of many forms can stall. Shipping the vague question costs a score you cannot explain to a supervisor. Split the work the way the admin guide already splits automation: - **Metrics** for numbers the contact already stores: longest hold, hold count, interaction duration, interruptions, sentiment score. - **Contact categories** when the phrase is a rule. The greeting example in the docs is "Thank you for calling" in the first 30 seconds, labeled `ProperGreeting`. - **Generative AI** for single-select and text questions a person could answer from the transcript plus the instructions. On fully automated submit, that path is capped at **10** questions per contact. The cap is the [existing conversational-analytics quota](https://docs.aws.amazon.com/connect/latest/adminguide/generative-ai-performance-evaluations.html), not a limit introduced on October 9. > **What broke** — Modeled on the October 9 sales-disclosure example, not a cited client. Structural validation passes. The warning asks for the exact disclosure text. The manager marks the finding resolved and activates. Auto-scores mark Yes when the transcript says "disclosure" and the required sentence is absent. Detection: a supervisor opens the contact next to the auto-score. Fix: put the required sentence in the instructions, choose **Save and validate** again, and activate only after that warning is gone. **Resolved** hides the row. It does not re-check the form. > **Reproduce this** — Run [`evaluation-form-validation-checklist.md`](https://www.factualminds.com/examples/architecture-blog-2026/connect-evaluation-form-validation/evaluation-form-validation-checklist.md). It covers the Region gate, the Create permission, one vague Gen AI question, one metric question, the rate-limit retry, and a check that Resolved did not re-run validation. --- ## How answers get filled at runtime Authoring checks and runtime scoring are different steps. Save and validate judges the form. After activation, conversational analytics fills answers on each contact. | Automation | Question types | Use it when | | --- | --- | --- | | Contact categories | Single select, multiple select | The expected words or call reason already exist as a rule | | Generative AI | Single select, text | The transcript and your instructions are enough to decide | | Metrics | Number | The fact is hold time, interruptions, sentiment, or a similar contact metric | Fully automated submission is a separate toggle, **Enable fully automated submission of evaluations**, plus a conversational-analytics rule that chooses which contacts to score. Evaluators can still edit and resubmit. Self-service contacts (a Lex bot or an AI agent) use the contact interaction type under **Additional settings**. They are a different target from the human-agent form, with the same authoring check once a question is on generative AI. ![Runtime flow for Connect evaluations: a customer reaches a human agent or Amazon Lex, and the evaluation form fills answers from contact categories, generative AI, or metrics, with an evaluator override and a separate auto-submit rule.](../../assets/images/blog/amazon-connect-evaluation-form-automated-checks-2026-02.webp) _Figure: Three ways a contact gets an answer, plus override and auto-submit. [Open the draw.io](https://www.factualminds.com/examples/architecture-blog-2026/connect-evaluation-form-validation/connect-evaluation-runtime.drawio)._ --- ## What to Do This Week 1. **Confirm** the instance Region is one of the **8**. Seoul and Cape Town can have Cases and still lack this check. 2. **Open** one live generative AI form and choose **Save and validate**. You need the Evaluation forms **Create** permission. 3. **Rewrite** warnings that fail transcript answerability, positive action framing, or external-system references. Put the exact sentence in the instructions. 4. **Validate again.** Do not treat **Resolved** as a second pass. If the side panel says you hit **3** in parallel or **30** in the hour, wait and retry. Structural validation still runs. 5. **Leave** fully automated submit off until a supervisor has compared auto-scores to recordings on the questions you rewrote. Need the form model, the security profiles, and the rule design done as one pass? FactualMinds is an AWS Select Tier Services Partner. Start at [Amazon Connect consulting](/services/amazon-connect/) or [contact us](/contact-us/). --- ## What This Post Doesn't Cover - **Connect Customer pricing** for voice, chat, conversational analytics, or evaluations. Use the current pricing page for your Region. - **A calibration study.** We have not scored a queue before and after these checks, and this post does not invent one. - **Scoring design.** Percentage versus points, automatic fail, weights, and performance thresholds are in Steps 5 and 8 of the same admin guide. - **PDF import** of a form from another quality tool (documented max **2 MB**). Useful, and out of scope here. - **The validation API.** Console path above is the one this post verified against the October 9 what's new and Step 9 of the admin guide. --- ## Related reading - [Amazon Connect Consulting](/services/amazon-connect/) - [Connect Cases CSV Export (Aug 2026)](/blog/amazon-connect-cases-csv-export-agent-workspace-2026/) - [HIPAA-Compliant Architecture on AWS](/blog/how-to-implement-hipaa-compliant-architecture-aws/) - [AWS for Retail: Complete Guide](/blog/aws-for-retail-complete-guide/) ## FAQ ### What did AWS ship for Connect evaluation forms on October 9, 2026? Save and validate on a performance evaluation form. Structural problems return as Errors and block activation. Generative AI questions are then checked against 11 best practices and return as Warnings. Warnings do not block activation. The Gen AI stage allows 3 validations in parallel and 30 per hour per instance. The feature is in 8 Regions. ### When should we not rely on this check? Skip generative AI for questions that need audio-only cues (volume, pitch, speaking speed, vocal tone) or a lookup in a system the transcript does not contain. Use metrics for numeric facts such as longest hold. Use contact categories when the expected phrase can be written as a rule. The check also does not run in Seoul or Cape Town. ### What could go wrong if we activate with warnings still open? AWS allows it. A vague question such as whether the agent gave a sales disclosure can score Yes because the transcript contains the word disclosure. Marking a finding resolved only hides that row. It does not re-run validation. Clear the warning, choose Save and validate again, then activate. ### Which Regions include the October 9 check? US East (N. Virginia), US West (Oregon), Canada (Central), Europe (Frankfurt), Europe (London), Asia Pacific (Singapore), Asia Pacific (Sydney), and Asia Pacific (Tokyo). That is 8 Regions. Cases CSV export also includes Seoul and Cape Town. An instance can have Cases and still lack this check. ### How many questions can generative AI answer on one contact? For fully automated submission, AWS documents a cap of 10 generative AI questions per contact. That quota is separate from the October 9 authoring check. Contact categories and metrics do not count toward the 10. Assist-the-evaluator mode is described separately in the generative AI evaluations guide. ### Should every contact be auto-submitted on day one? Leave fully automated submission off until a supervisor has compared auto-scores to recordings on the questions you rewrote. The toggle and the conversational-analytics rule are a second decision, after Save and validate. Evaluators can still override an automated answer before they submit. --- *Source: https://www.factualminds.com/blog/amazon-connect-evaluation-form-automated-checks-2026/*