⚙️ For client admins
Self-registration and seat limits now work together. For admins with cohort seat limits specified, those limits now automatically disable self-registration options for the cohort once the limit has been reached. No manual intervention is needed to stop a cohort from overfilling. On a related note, the self-enrollment option is now enabled by default.
For organizations running in-person sessions, sessions can now be configured to disable the automatic prompt to join video, which removes a confusing step for groups already in the same room. Contact us to request this for your account.
✨ For session participants
Participants viewing a screen-share from another participant will now see a new full screen option to enlarge the shared screen — useful when the shared content is dense, or the detail that matters is small.
🏛️ For module authors
Architect structures module development as a series of phases, moving from rough initial idea to participant-ready module, with authors deciding when and how AI does the work. This month we made the early phases far more forgiving and the later phases far more transparent.
Start rough, then find the structure. Vision no longer needs to arrive fully formed. Authors can open Phase 1 as a blank canvas, jot down rough notes in whatever order they come, and then use Refine Vision to have Archie recover the structure — asking a few plain-language questions where the notes are ambiguous — before moving on to Phase 2.
Draft activities from a plan, not one prompt at a time. Archie now drafts Activities one activity at a time, working from a design brief that allocates scenarios and time across the whole module. That means each activity is drafted with the rest of the module in view, rather than in isolation. Auto-drafted outlines also now follow the pedagogical sequencing we hold our own modules to: every scenario, framework, or term is introduced on a read-only step before any step asks participants to use it, with one objective per activity.
Describe resources, then have them written. While still outlining activities in the Activities step, authors can describe what each step's resources should contain, then have Archie auto-draft the resource content from that description. Archie-drafted step content now also comes back in one consistent Sparkwise writing voice, so a module reads as a single piece of work.
Give AI chatbots a role to play. Authors can now add custom system prompts to the AI chatbots inside a module, so a chatbot can support a specific scenario or take on a specific persona — a skeptical stakeholder, a customer with a complaint, a coach — rather than acting as a general-purpose assistant.
See what Archie did, and undo it. The resource drawer now shows the exact prompt that produced each resource, so authors can see how a draft came about and adjust the input rather than rewriting the output. And the Activity log now surfaces every edit Archie makes with built-on-the-fly tools, with one-click revert for any of them.
Check a whole phase, then polish. Authors can evaluate an entire phase in one click: Archie re-runs every check, narrates the results, and, when everything passes, offers to advance to the next phase and start its draft. Modules now also finish in a dedicated Polish phase, focused specifically on publication-readiness.
Architect is currently available to clients working with us on custom module builds. Reach out if you'd like to see it in action.
That's the update for this month!
Interested to see what we've released in past months? Highlights from our changelog can be found in our Product Release Notes.
