Roll out every policy change in days, not weeks.

Juno finds everyone a new policy applies to, updates the training, collects each acknowledgment and shows who is done, so nothing slips before the audit.

Example data

Global-eWalkMeCellebriteeToroJFrogGongCintJobber

A new policy is out. Now everyone has to read it.

Emails, reminders and a spreadsheet of who signed: every policy change turns into weeks of chasing.

How it works

1Trigger & Signal

  • Google Drive / SharePointLegal's new version, the diff

2Knowledge Design

Your LLM + Juno MCP

  • Reads the diff
  • Scopes who the clause binds
  • Updates last year's training
You sign the scope off

3Delivery & Adoption

4Measure

Days from publication to current, on this version

Tuesday, 4 pm: the new version, rolled out in one conversation

Legal published data-retention v3. Roll it out to everyone it binds, with a check and an acknowledgment.

Claude

Example data

See the version each person consumed, beside what they scored

Example data

  • Version Consumed

    Each learner's record names the published version of the course they saw.

  • Beside the score

    The quiz result and its attempts sit on the same row as the version.

  • The instrument, named

    A signature, a certification or a pass: the record says which one.

  • Filter, save, export

    Filter to one policy, save the view and export it as a CSV.

Narrow it to exactly the people the clause applies to

Example data

  • Criteria as nouns

    Role, position, department, location, group and employee type decide who is bound.

  • Contractors, by type

    Contractors are an employee type, so the rule scopes them in or out.

  • A dry run first

    A dry run counts the people a rule will touch before anything sends.

  • Your sign-off

    The compliance owner signs the scope off before anyone is assigned.

For the people who join after this rollout: Company & Dpt-Wide Orientations

Start the rollout from the change itself

Example data

  • Your LLM reads the diff

    Your model reads what changed between the two versions of the document.

  • Three priorities

    Mission-critical, Important, not urgent, Nice to know. The change lands in one.

  • Where your policy lives

    Confluence, Notion, SharePoint, the GRC tool or the audit tracker, reached by your LLM.

  • Updated in place

    Last year's course is updated as a draft, then published as the new version.

Juno acts within your existing Juno permissions, over OAuth, with no API keys.

The same loop runs on a product release: Release & Product Update

See securitySee Juno in your LLM

Nobody keeps a chase list

Example data

  • Due dates that notify

    Notifications go out 7 days before, 1 day before and 2 days after.

  • Nudge and Notify

    Two automation actions, on a rule you set, for whoever has not started.

  • A manager's own list

    Each manager sees who on their team is overdue.

  • Certifications that expire

    A certification you created carries its expiry date and its reminders.

  • Done means done

    Reminders stop for each person once their work is complete.

From Tuesday to the 30th, on the calendar

  1. Tue4 pm

    Legal publishes v3
  2. Wed

    Scope signed off

    1,412 people

  3. Wed

    Assigned
  4. Day 7

    Reminders to those not started
  5. Day 14

    88 % acknowledged v3
  6. Day 30

    Day 30

    • Current: 1,389
    • Exceptions signed off: 23

Example data

Seasons planned in advance

These teams planned their compliance season before it started. The completion figures are theirs, in their own words.

“Running our compliance season with Juno was a huge success. With everything planned out in advance from automated assignments and reminders to real-time analytics, we were able to drive 100% completion across the org.”
Rotem W.Senior Global L&D Partner, Cloudinary
“…smooth, structured, and highly effective… over 90% completion across the company, and the process felt completely under control from start to finish.”
Marion D.People Experience Partner, Efficy

Trained thousands on key topics including compliance, information security, financial monitoring and personal safety.

85–98% completion on mandatory programs

76% of employees rated Juno 9 or 10

Raiffeisen Bank case studySeptember 2025

How fast the organization actually turned

The number to watch is not the share who acknowledged. It is the time between publication and a population current on this version.

Example data

Try it

  • Roll out a new policy with proof of completion

    A short course, a 5-question check and an acknowledgment, with a clear list of who's done.

    Your AI asks you

    The policy document, who must complete it, the deadline, acknowledgment or quiz

    The prompt

    I want to roll out a new policy in Juno Journey with proof that everyone who needs to has read and understood it. Use the Juno Journey connector for anything in Juno. If it isn’t connected, stop and tell me how to add it (https://docs.junojourney.com/mcp/overview). Before you build anything, ask me: 1. The policy document (upload it, or tell me where it lives) 2. Who must complete it (departments, locations, roles) 3. The deadline 4. Acknowledgment, a quiz, or both Ask these in one message, then wait for my answers. If the document lives in a tool outside Juno, check it’s connected in this chat; if not, I’ll upload it. Stick to what the policy says; don’t add rules. Search Juno for older versions of this policy training and update rather than duplicate. Then plan a short course (what changed, what it means for you, examples), a 5-question check with a pass mark, and an acknowledgment step. Show me that outline, the audience and the due date first, then build as drafts. Don’t publish, assign or notify anyone until I explicitly approve.

  • A completion heatmap by team and topic

    Teams down the side, topics across the top, so the gaps show up before the deadline does.

    Your AI asks you

    Which groups and topics, which courses count, what "done" means, the date range

    The prompt

    I want a completion heatmap: groups down the side, topics across the top, and each cell showing how far along that group is. Use the Juno Journey connector for anything in Juno. If it isn’t connected, stop and tell me how to add it (https://docs.junojourney.com/mcp/overview). Before you build anything, ask me: 1. Which groups (departments, roles, locations) 2. Which topics, and which courses count for each 3. What “done” means (completed, passed, above a score) 4. The date range Ask these in one message, then wait for my answers. Pull everything you can from Juno; if you can’t read completions here (analytics may be off for our account), I’ll upload an export from Juno. If a grouping lives outside Juno (e.g. an HR system, if we use one), check it’s connected in this chat; if not, I’ll upload a CSV. Never guess anyone’s group. Confirm definitions with me, then build an interactive artifact: the heatmap with counts per cell, a click-through to the people in each cell, and the three weakest spots with suggested next steps. State definitions, show sample sizes, and grey out cells too small to judge. Don’t assign or notify anyone.

  • Find training that's out of date

    Claude checks your courses against your source of truth and lists what to fix first.

    Your AI asks you

    Where your source of truth lives, which topics matter most, what counts as a recent change

    The prompt

    I want to find training in Juno Journey that no longer matches how things actually work today. Use the Juno Journey connector for anything in Juno. If it isn’t connected, stop and tell me how to add it (https://docs.junojourney.com/mcp/overview). Before you build anything, ask me: 1. Where our source of truth lives (product docs, release notes, policies) 2. Which topics or courses matter most 3. How recent a change must be to count Ask these in one message, then wait for my answers. Check that source is connected in this chat; if not, I’ll connect it or upload the documents. Only flag drift you can show evidence for. Read the relevant Juno courses, quizzes and journeys, compare each claim to the source, and give me a table: course, what it says, what the source says now, severity, suggested fix. This is a review, not an edit: change nothing yet. Offer to draft updated versions of the top items; after I approve, save them as drafts and don’t publish until I say so.

Every prompt

Your resellers and your customers are bound by documents too

  • PartnersA partner who has not re-attested is exposure with a contract attached.
  • CustomersCustomer users acknowledge the new terms before they go further.

Questions people ask.

How do we roll out a new policy and make sure everyone acknowledged it?

Start from the change. Your LLM reads the new version and drives Juno: it finds the people the policy binds, updates last year's training and prepares the assignment with a due date. You sign the scope off before anyone is assigned. Each person then acknowledges the version they saw, and you see who is done.

Which version of the policy did each employee acknowledge?

Juno records it. Training Analytics carries a Version Consumed field: the published version of the course or quiz each learner saw. It sits on the same record as their score, their attempts and the date, so you can filter to one policy, see who acknowledged the current version and who is still on the last one, and export it.

How do you know employees understood a policy, beyond receiving it?

You set a check with a real bar. A knowledge check carries a pass mark, a number of attempts and a time limit, and open answers can be graded by a validator. Time spent counts engagement only. The result sits beside the version the person consumed, so the record shows what they answered about the change itself.

Do we need policy management software or a learning platform?

They do different jobs. Policy management software holds the document of record: its versions, approvals and history. Juno is not your policy repository. It makes the people a policy binds current on the new version, captures each acknowledgment against that version and shows who understood it. Many teams run both, and your LLM can read the policy wherever it lives.

What counts as a policy acknowledgment: a signature, a certification or a quiz?

All three are available, and you choose. A signature element captures the person's signature inside the training. A certification you design is awarded by rule. A knowledge check records a pass against a bar. Whichever you pick, the record names the person, the version they consumed, the date and the instrument. Your counsel decides what your policy requires.

How do we handle a policy that only applies in some countries or to some roles?

You scope it with criteria: role, position, department, location, group and employee type. A dry run shows how many people the rule will touch before anything is sent, and the compliance owner signs the scope off. When one policy has a different version per state or country, see HR & Workplace Conduct.

Can our LLM read our policy in Confluence or SharePoint and assign the training?

Yes, when your LLM is connected to both. It reads the policy where it lives, through that tool's own MCP connection, and drives Juno through Juno's MCP server: it finds the existing training, drafts the update and prepares the assignment. Juno does not watch your documents. Your LLM starts the work, and you approve it.

Does an AI decide who has to take compliance training?

No. The system executes; people approve. Your LLM proposes the scope, and the compliance owner signs it off. Publishing, assigning and notifying each need an explicit review before they are confirmed, and drafts stay invisible to learners until published. Juno acts within your existing Juno permissions. How we handle data is on Security.