This release brings AI-powered practice with Roleplays to your programs. Users often walk into high-stakes conversations unprepared, and even when they are building a skill with a mentor, they rarely get a safe place to try it before it counts. With Roleplays, users can practice any conversation your organization sets up in private and get feedback on how they did.
This release includes:
Roleplay practice: users talk with a realistic AI persona, by voice or text
Readiness feedback: a private debrief after every practice
Roleplay authoring: admins build roleplays, personas, and objectives in Admin Hub
Chronus Roleplay Library: ready-made roleplays admins can import and adapt
Access and reporting: control over who practices what, with aggregate metrics
Together they form a loop: admins publish scenarios, users practice and get feedback, and they come back until they feel ready for the real conversation.
Applicable For
Feature | Who it's for | Enablement |
Roleplay practice and readiness feedback | Users (also includes administrators) | Enabled for the organization or for specific programs |
Roleplay authoring and the Chronus Roleplay Library | Organization and program administrators | Available in Admin Hub once Roleplays is enabled |
Please note:
Roleplays is a separate add-on. Please contact Chronus support or reach out to your Chronus account team and we'll help you get started.
Roleplays can be enabled for the whole organization or for specific programs, which suits a pilot with a limited group.
How it works
Roleplay practice for users
Once a roleplay is available to them, users see a Roleplays item in the navigation and roleplay cards on their home page. They can also find roleplays under Discover → Roleplays. The first visit explains what a roleplay is and that practice is private by default.
From there, users can:
Discover the roleplays available to them, ordered by relevance to their profile and goals they are working on.
Open a roleplay to see the scenario, their role and goal, the persona they will talk to, and how they will be evaluated.
Practice by voice or text. The persona opens the conversation and stays in character. It pushes back according to its demeanor, reveals what it knows only as the user draws it out, and steers off-topic conversation back to the scenario.
Practice again as often as they like. Results are kept per attempt so users can see their progress.
Readiness feedback
When a user ends a session, Chronus evaluates the conversation against the roleplay's objectives and shows:
A readiness result: Keep Practicing, Almost There, or Ready. It is a simple read on whether the user is ready for the real conversation, not a percentage or a grade.
Key highlights: what went well and what to work on, each tied to something the user actually said, with a concrete suggestion to try next time.
A result for each objective, with short feedback and a quote from the conversation as evidence.
Delivery insights for voice sessions: pacing, word choice, and brevity, and whether each improved since the last attempt.
The full transcript.
Users can rate the feedback with a thumbs up or down, which helps us keep improving its quality.
Organizations can turn off the readiness result. Users then still get highlights, objective results, and the transcript, with encouraging copy in place of the readiness result.
Sharing: results are private by default. Users can choose to share them with their mentors or with anyone in their organization. The person they share with is notified and sees a read-only view of the results. Users can stop sharing at any time.
Roleplay authoring for admins
Admins manage roleplays in Admin Hub → Roleplays, which has three tabs: Roleplays, Personas, and Objectives.
The roleplay builder covers the scenario and description, the user's role and goal, voice and/or text modes, the persona and its scenario-specific context, the objectives the roleplay is evaluated on, and who can access it.
Personas are reusable AI characters, each with a name, role, company, background, demeanor (Friendly, Neutral, Challenging, or Adversarial), a voice, and an avatar. One persona, such as a skeptical CFO, can be reused across many roleplays.
Objectives define what good looks like and drive the feedback:
Skills: rated on a five-level scale the admin describes
Behaviors: whether a behavior was demonstrated
Frameworks: multi-part methods, such as SBI or STAR, evaluated part by part
Qualitative: written feedback only, with no score
Preview lets admins run any roleplay, including a draft, exactly as a user would, by voice or text, before publishing.
Program administrators can create and manage roleplays for the programs they administer. Organization-level roleplays stay managed by organization administrators.
Chronus Roleplay Library
Admins can start from a library of ready-made, Chronus-authored roleplays instead of building from scratch. They can browse and import the ones they need. Each import arrives as a draft, with its personas and objectives, so nothing reaches users until an admin publishes it, and every copy can be edited to match the organization's language.
Access and reporting
Who can access: a roleplay can be available to the whole organization or narrowed to specific programs and audiences.
Reporting: admins see aggregate metrics, such as how many users practiced, the number of attempts, and practice time, across the organization and for each roleplay. Program administrators see the same metrics, scoped to their own programs.
Why This Matters
Confidence before the conversations that count. Users rehearse the conversations that matter to them somewhere safe, and walk in prepared.
Skills that grow between mentoring sessions. Users practice the skill they are working on with their mentor, share results, and see it improve, attempt by attempt.
Feedback that points to the next step. Highlights and objective results show users exactly what to work on, and the readiness result tells them when they are ready.
Fast time to value for admins. With the Chronus Roleplay Library, admins can launch relevant, organization-specific practice without help from Chronus.
Compliance & Security
Roleplays were built with privacy and customer control in mind:
Practice is private. Individual transcripts and results are visible only to the user and to the people they choose to share with. Admins see aggregate metrics only, never an individual user's conversation or result.
Customer data is never used for model training, and there is no cross-tenant learning.
AI is clearly disclosed. Users and admins see clear notices that roleplays use AI and that results are AI-generated.
Subprocessor and transparency updates accompany this rollout. The Chronus Subprocessors page and the AI @ Chronus transparency page are updated for Roleplays.
For the latest details, see the Chronus Subprocessors page and the AI @ Chronus transparency page in the Help Center.
Release Certification
This release has undergone rigorous testing phases to ensure its reliability and efficacy. We are proud to report that the release has been thoroughly tested and certified by Chronus. This certification signifies our unwavering commitment to maintaining the highest standards of quality for our users.