Frequently asked questions

No. Your uploaded documents, prompts, and the content you create on Creatium are never used to train any AI models, including the third-party LLMs our platform uses under the hood. We have strict data processing agreements in place that prohibit this. Your data stays your data.

Absolutely not. Every organization on Creatium operates within a fully isolated environment. The content you build, the knowledge bases you upload, and the conversations your learners have are gated entirely to your organization. Nothing leaks across accounts.

You do. All content you create on the platform (videos, roleplays, simulations, coaches) is fully owned by your organization. You can export it, embed it, and distribute it however you choose.

Published content that has been exported as a SCORM file or embedded via link/iframe will continue to work independently. However, interactive features like AI roleplays and coaching conversations require an active Creatium account, as those run on our AI infrastructure in real time.

We work with organizations that have some of the most rigorous data security requirements in the world, including top 3 management consulting firms, the largest school districts in the United States, and multiple Fortune 100 financial institutions. All of our data handling practices have been reviewed and approved through their enterprise procurement processes.

Creatium works with virtually any LMS that supports standard integration methods. You can export content as a SCORM file (compatible with platforms like Canvas, Docebo, D2L, UKG, Sum Total, Cornerstone, and any SCORM-compliant LMS), embed it as an iframe, or distribute it as a simple shareable link. If your team uses authoring tools like Articulate Storyline or Rise, you can also embed Creatium content within those courses.

Creatium content works on mobile phones, tablets, and desktop computers. The platform supports full keyboard and mouse navigation for accessibility. We’ve completed extensive accessibility reviews with enterprise partners to ensure compliance with accessibility standards.

Yes. Every roleplay and coaching conversation supports both voice and text input. Learners can speak naturally using their microphone or type their responses, whichever they prefer. This also makes the platform accessible for environments where audio isn’t practical, like an open office or a library.

Yes. The interactive video editor includes an Embeds panel where you can insert MP4 video files, YouTube links, PDF documents, Excel files, and external URLs directly into a lesson. Embedded content appears within the Creatium video player, so learners never leave the learning experience.

Yes. When creating a new lesson, you can select the output language during the content generation step. The platform will generate slides, narration scripts, and learning objectives in the selected language.

Creatium uses a combination of proprietary and third-party AI models, each optimized for different parts of the platform. We’ve built our own specialized LLMs for Socratic tutoring and conversational coaching, which outperform general-purpose models like ChatGPT for learning use cases. We also use best-in-class models for image generation, voice synthesis, and content creation, and we update these as better options become available. Critically, none of your content is ever used to train any of these models.

Yes. Creatium supports full brand customization. You can apply your organization’s color scheme, upload logo assets, and use brand-aligned slide templates across all content. For enterprise partnerships, we also do custom prompt development on the back end so that every piece of generated content (from slide layouts to image styles) aligns with your brand guidelines automatically.

Yes. You can create a custom avatar from a headshot photo and a short voice sample. The platform generates a lifelike animated presenter that uses your team member’s appearance and voice. We also support animated character avatars (like a mascot or illustrated figure) if that’s a better fit for your brand. The individual will need to provide consent through a simple sign-off process.

Yes. On any slide, you can upload your own images to replace the AI-generated options. You can also upload custom background images for roleplays (for example, a photo of your office lobby, a pharmacy counter, or an airplane aisle) to make the experience feel more realistic and on-brand.

PowerPoint animations do not transfer when you upload a slide deck to Creatium. The platform captures each slide as a static image. If your slides rely on animations (like content that builds in step by step), we recommend converting each animation state into its own separate slide before uploading. Alternatively, you can screen-record the animated slide playing in PowerPoint presentation mode and embed that recording as an MP4 video within your Creatium lesson.

The platform does not currently support standalone audio file embeds. However, you can work around this by screen-recording the audio playing over a slide (capturing it as a video), and then embedding that video file into your lesson. Native audio embed support is on our product roadmap.

This is a feature currently in development. Today, each AI coach has its own individual knowledge base. The ability to create a shared, organization-wide knowledge base (where you can upload content once and have multiple coaches, roleplays, and team members reference it) is one of our most-requested features and is actively being built.

Extremely detailed. You can define the character’s emotional state, personality traits (including Myers-Briggs types), escalation triggers, specific phrases they use, how they react to different responses, and whether they de-escalate or become more confrontational based on the learner’s approach. You can also define the opening line, conversation duration, and specific knowledge the character should reference. Everything compiles into a master prompt that you can directly edit for even finer control.

Creatium uses a rubric-based evaluation system. When you create a roleplay, the platform auto-generates a scoring rubric based on the scenario you’ve designed. The rubric evaluates across categories like information gathering, empathy, communication style, and problem resolution. You can define mastery phrases (what good responses sound like) and pitfalls (responses that should be flagged). After the conversation, the learner’s transcript is scored against this rubric and an evaluation report is available for download.

Yes. The analytics dashboard captures complete transcripts for every coaching and roleplay session. Administrators can view these at the individual learner level, review specific conversations, and track patterns across your team. This data is invaluable for identifying coaching opportunities and understanding where learners are struggling.

Creatium tracks completion metrics, watch time, number of attempts, activity scores, and roleplay evaluation results at the individual learner level. You can see who completed a training, how long they spent, how they performed on activities and assessments, and review their conversation transcripts. This data also flows into your LMS for centralized learner records.

Yes. You have several options for access control. You can restrict content to specific email domains (for example, only users with a @yourcompany.com address can access it). For enterprise accounts, you can create learner groups to segment access by team, department, or role. You can also revoke access to any published content at any time, useful when training materials are updated or a module is no longer relevant.

Not currently. The AI evaluates learner performance based on the content and quality of their spoken or typed responses. You can enable the learner’s webcam during a roleplay so they can see themselves on screen (which adds realism and self-awareness), but the AI does not analyze visual input. Voice and text analysis are the current evaluation modalities.

They serve different purposes in the learning journey. An AI Coach is the expert: the learner comes to the coach with questions, and the coach guides them using Socratic tutoring and the knowledge base you’ve uploaded. Think of it as an always-available trainer or subject matter expert. An AI Roleplay flips the dynamic: the learner becomes the expert, and the AI plays a character (a frustrated customer, a confused patient, a skeptical prospect). The learner must respond in real time and is evaluated on how they handle the conversation. Think of it as a practical assessment or oral exam. Most organizations use coaches for learning and reinforcement, and roleplays for practice and assessment.

Yes. Once you create a coach or roleplay, it lives in your library and can be inserted into as many lessons as you like. You can also share it as a standalone link for use outside of video-based lessons: for example, embedding a coaching link in an email, on an intranet page, or within an existing course on another platform.

No. Changes to roleplay characters and coaches update in real time across all published content. If you tweak a character’s personality, add new knowledge base documents, or adjust the evaluation rubric, those changes take effect immediately for all learners, no republishing required.

This is a valid concern and one we take seriously. The most effective safeguard is to explicitly restrict the coach’s prompt to only reference the uploaded knowledge base. In the master prompt, you can add a clear instruction like: “Do not reference any information outside of the uploaded knowledge base. Restrict all responses to the documents provided.” This ensures the coach draws only from your approved materials. We also recommend testing the coach with challenging questions before deploying it to learners, and iterating on the knowledge base and prompt until the responses consistently meet your standards.

Yes. The AI coaching system adapts to the learner’s demonstrated knowledge level during the conversation. If a learner is clearly expert-level, the coach can skip foundational content and move directly to advanced scenarios or case studies. You can also configure this behavior in the prompt: for example, instructing the coach to assess the learner’s baseline knowledge in the first few exchanges and adjust depth accordingly.

Yes. Creatium’s Video Viewer (the interface learners interact with when consuming content) has been evaluated against WCAG 2.0, 2.1, and 2.2 at both Level A and Level AA, and supports conformance across all applicable criteria. The platform is fully operable via keyboard-only navigation with no keyboard traps, all interactive controls expose proper name, role, and value information for screen readers through ARIA attributes, and visible focus indicators are present on every interactive element. We maintain a formal VPAT (Voluntary Product Accessibility Template) Accessibility Conformance Report, which is available upon request.

Yes. Every roleplay and coaching conversation on Creatium supports both voice and text input. Learners can speak naturally using their microphone or type their responses, whichever they prefer or whichever their environment requires. The entire platform is navigable with a mouse, keyboard, or touch input.

Yes. Creatium automatically generates subtitles for all video content created on the platform. Captions are available in multiple languages and can be selected by the learner directly from the video player’s interface menu. This ensures that all narrated training content is accessible to learners who are deaf or hard of hearing, or who simply prefer to read along.

Yes. Every AI coaching session and roleplay conversation generates a complete transcript that is stored in the analytics dashboard. Administrators can review these transcripts at the individual learner level to track performance, and learners receive feedback summaries after each session. This means that even though coaching and roleplay interactions are primarily voice-driven, the full content of every conversation is captured in text form for review, record-keeping, and accessibility purposes.

Creatium content works on mobile phones, tablets, and desktop computers. The Video Viewer supports both portrait and landscape orientations without locking or restricting screen rotation. We’ve completed extensive accessibility reviews with enterprise partners to ensure compliance with accessibility standards across a wide range of devices and environments.

Creatium’s accessibility conformance covers the core Video Viewer and its standard controls, the interface your learners interact with directly. We recommend that organizations review any custom or third-party content they embed for accessibility compliance in the context of their own requirements.