Knowledge intake takes a whole folder, the ideas atlas reads truthfully, and chat reasons from more of your evidence.
Product-knowledge intake can now process an entire folder of documents in one run — using each file's own outline for structure — and it stays honest and recovers when the AI service is slow or unavailable. The Discovery ideas atlas treats Size as a visual control, shows every idea (unscored ones included, in edge lanes), and large opportunity trees open again at scale. Workspace chat and connected agents can query your signals and idea votes, and the assistant answers faster and holds its state.
What changed
- Process a whole folder of product documents in one run, with sections from each file's own outline
- Knowledge intake handles AI outages honestly — stuck runs recover, and files are never marked unreadable on a bad service call
- The ideas atlas shows every idea: count-based views include zero counts, and unscored ideas appear in labeled edge lanes
- Workspace chat and agents can query your signals and answer which ideas have the most vote demand
- Large opportunity trees open reliably again
- Workspace chat responds faster
Bring in product knowledge a folder at a time
Knowledge intake now processes a whole folder in one run, building sections from each document's own outline and keeping a digest of what came in. When the AI is slow or unavailable, runs say so and offer a retry — files are never marked unreadable because of a service outage, and stuck runs recover on their own.
- Import a folder of documents in one run with sections that follow each file's structure
- AI outages are reported as retriable, not as file errors
- Stuck runs recover without manual intervention
The ideas atlas shows every idea
Count-based axes used to hide any idea with a zero count, and changing Size or clicking an idea triggered a full refetch. Now the atlas shows every idea — zeros included — Size resizes what's loaded, and ideas without scores appear in labeled edge lanes instead of being excluded.
- Votes, linked insights, and ARR axes actually populate
- Size and axis changes feel instant
- Unscored ideas appear in edge lanes instead of vanishing
Chat reasons from more of your evidence
Workspace chat and connected agents can now query signals by topic, source, account, product, or status, and answer questions about which ideas have the most vote demand. Responses are faster.
- Ask about signals and get answers tied to their sources and accounts
- Ask which ideas have the most votes, or which accounts are asking
- Faster answers with the same curated context
More in this release
Discovery and product knowledge
Folder-scale knowledge intake, a truthful ideas atlas, and Discovery at scale.
- Knowledge intake processes a whole folder in one run, with sections from each file's outline
- Knowledge intake handles AI outages as retriable and recovers stuck runs
- Ideas atlas: count-based axes include zero counts; Size and selection don't refetch
- Unscored ideas appear in labeled edge lanes instead of being hidden
- Opportunity trees load from bounded counts instead of a deep relation chain
- Idea trace layout and saves stabilized
- Discovery section view dropdowns refined
Chat, agents, and the app
More of your evidence is reachable from chat and connected agents.
- Workspace chat and MCP can query signals by topic, source, account, product, or status
- Chat answers vote-demand questions — top-voted ideas and voting accounts
- Workspace chat responds faster
- Workspace navigation stabilized: header holds state, initiative routing fixed
- Jira import shows only the fields that need your attention
- Settings Team navigation aligned
- Harvest voice input steadier