Chat now shows open agent questions, live activity, and research status in one panel, competitive landscape and Target Intelligence results are more accurate, and deep research exports as editable branded slide decks.
This release makes research chats easier to follow while they run, improves the accuracy of competitive and target evidence, and gives you editable, branded slide decks from deep research.
New: Question Panel and Live Activity in Chat
When an agent needs your input, its questions now appear in a structured panel you can answer and pick up later, alongside a live activity list and a research status card that shows where a run stands.
New: Editable Branded Slide Decks
Deep research reports now export to PowerPoint with a branded layout you can edit directly, instead of a fixed-format presentation.
Improved: Competitive Landscape Evidence
Asset evidence is now built from real Target Intelligence data, and citations and clinical trial identifiers are more accurate.
Improved: Antibody Reports
Antibody agent reports follow a clearer template, and the steps of a long run are grouped so its progress is easier to follow.
Fixed
Target Intelligence no longer shows “Unknown” as a trial’s drug name, antibody agent reports load reliably, and we resolved overlapping labels and hard-to-read headers in the Score Breakdown and Report panels.
And as always
We’ve kept expanding our antibody and macrocycle design pipelines behind the scenes, so the candidate sets they produce keep getting richer.