> ## Documentation Index > Fetch the complete documentation index at: https://docs.wisdom.ai/llms.txt > Use this file to discover all available pages before exploring further. # Changelog **Updates** ### ⚡ Wisdom Smart Cache Apps and dashboards no longer have to query your warehouse on every load. Wisdom Smart Cache serves them from Wisdom-managed snapshots of your tables, rebuilt on a schedule you control. Any query the cache can't serve runs against the warehouse, so caching changes how fast users get answers, never the answers themselves. Row-level security and column masking still apply at query time. For full details, see [Cache tables for faster apps and dashboards](https://docs.wisdom.ai/setting-up-wisdom-ai/cache-tables.md) and the [Smart Cache screen](https://docs.wisdom.ai/setting-up-wisdom-ai/smart-cache.md) pages. **⚙️ How it works:** * **Register tables for caching:** In **Domains → Data Sources**, register every cacheable table in a database, or a single table from its **...** menu. Choose a rebuild interval (hourly, 3, 6, 12, or 24 hours), add sort columns, or select **Build now** * **Smart Cache screen:** Open `/caching` to see the share of queries served from cache, the speedup, total data cached, and caches that need attention. **Candidates** ranks every app and dashboard by the warehouse time caching could win back * **Optimize:** Select **Optimize** on a candidate and WisdomAI registers the table caches the app's widgets are missing. Where a table is too large to cache, it creates cached views shaped to those widgets. Your app is never edited * **Activity tab:** See every cached table and view with its status, freshness, size, and queries served. Configure it, review its builds and usage, trigger **Build now**, or **Stop caching** * **Cached data on dashboards:** When a tile is served from the cache, the header shows **Data as of** the oldest data on the artifact. See [Cached data on dashboards and apps](https://docs.wisdom.ai/dashboards/cached-data.md) > [accordion] FAQs > > > [accordion] Does a table rebuild every time its interval elapses? > > > > A table rebuilds when its interval has elapsed and its source data changed. If the warehouse reports no change, the existing snapshot is kept at no warehouse cost. **Build now** always forces a rebuild. > > > [accordion] Can I turn caching off? > > > > Yes. Go to **Domains → Advanced settings → Performance** and turn off **Materialized table cache**. Registrations are kept, and queries simply stop reading from the cache. > > > [accordion] Does refreshing a dashboard rebuild the cache? > > > > No. **Refresh Data** re-runs every widget query. Cached tables read the latest snapshot and everything else hits the warehouse. Rebuilds are controlled from Domains or with **Build now**. > [!NOTE] > > Wisdom Smart Cache is in beta and is enabled per account. The Smart Cache screen is available to superusers. ### 🪞 Object Mirroring Share domains and dashboards across customer organizations without maintaining a copy for each customer. Your team keeps the originals in an authoring organization, and each customer uses a live, read-only mirror in their own private organization. Changes to the original appear in every mirror. For full details, see [Object mirroring](https://docs.wisdom.ai/access-management/object-mirroring.md). **⚙️ How it works:** * **Create a customer organization:** A Site Administrator creates a separate organization per customer, and you invite that customer's users there. They don't need membership in the authoring organization * **Mirror a domain or dashboard:** Select the **Mirror domain** icon on a domain, or **Mirror dashboard** from a dashboard's **More** menu, then choose the destination organizations. Domains and dashboards are mirrored separately * **Choose target access:** **Admins only** lets administrators review a mirror before rollout. **All users** gives every user in the selected organizations read access, including users added later * **Customer experience:** Users can view mirrored dashboards, browse mirrored domains, and chat using them. They can't edit the shared content or see another customer's organization * **Data access still applies:** Row-level security, column-level security, and connection parameters follow each viewer's own identity. Source chats and authoring metadata stay hidden from customers * **Automate with the API:** Provision organizations, users, and mirrors with GraphQL. See [Set up tenant organizations and mirrors with the API](https://docs.wisdom.ai/integrations/user-management/mirror-with-the-api.md) > [accordion] FAQs > > > [accordion] Do customers get a copy of the domain or dashboard? > > > > No. A mirror refers to the original, so it stays synchronized with it. Edit the original in the authoring organization and you don't need to republish for each customer. > > > [accordion] Does mirroring a dashboard also mirror its domain? > > > > No. Mirror each object separately. You also can't mirror an existing mirror into another organization. Create each mirror from the original. > > > [accordion] Can I mirror to more than 100 organizations? > > > > The mirror dialog loads up to 100 eligible organizations. For larger rollouts, provision mirrors through the API. > [!NOTE] > > If you don't see **Mirror**, ask your administrator to check your permissions and whether mirroring is enabled for your account. In the API, this feature still uses the names publish and publication. ### 🧩 Embed Live Apps Embed WisdomAI Live Apps directly in your own application with an iframe, with no separate login for your users. For full details, see [Embed a Live App](https://docs.wisdom.ai/integrations/embeddings/iframe/embed-a-live-app.md). **⚙️ How it works:** 1. Generate a JWT for the user, as described in the [Embeddings Overview](https://docs.wisdom.ai/integrations/embeddings/embedding.md) 2. Insert `/embed/` before the app's name and ID in the app URL, and append the JWT as a `token` query parameter: `https://.wisdom.ai/apps/embed/name-id?token=` 3. Use that URL as the `src` of an `