Feature Requests

Honestly, the auto tag feature kind of lacks, to be generous. For example I can take a video with a persons name in the title or the channel name has a tag created but the auto tag will only give one tag “health,” “business,” etc. There could easily be 5 additional tags already created that would go with the title alone, not to mention, connections which have tags. As a sort of paraphrased example “Recipes,” “GordonRamsay,” “HowTo,” “PastaSauce.” Whatever. The singular auto tag is “Health.”It could even give “suggested tags” but instead I have to wait forever to process then tag it. Really, if it’s only giving 1 auto tag, why give any? It usually doesn’t even pick the most logical tag. And why can’t we set a default tags for every that comes in like a Zettelkasten type system where I want to tag all “ToProcess” or “ToBeCategorized” and once a week sit down to review and do whatever with them.

have a notebook that can be opened alongside the video summary. I know there is already a notebook for each individual video, but what I would personally find even more useful is one general notebook where I could add notes from different videos and keep it open next to the video summaries.

As a Recall user, I want advanced AI-driven recommendations for tags and links when creating or updating knowledge cards, So that I can enrich my knowledge base with more accurate connections and categorizations, improving the discoverability and utility of the information I collect as it continues to grow in size, reducing the amount of manual updates and links needed. Acceptance Criteria Inferred Relationships Discovery: The system automatically suggests potential relationships between knowledge cards based on the content, existing links, and tags within the knowledge base, when a card is created or upon user request for existing cards. Dynamic Tag Recommendations: When a new knowledge card is created or during review of an existing card, the system suggests relevant tags based on the card's content and the context of existing tags within the knowledge base, enhancing content categorization. Link Recommendations: Upon creation of a new card or through user initiation on an existing card, the system suggests potential links to other knowledge cards that share related content or concepts, facilitating deeper connections within the knowledge base. Actionable Suggestions: Users receive actionable suggestions that they can accept, modify, or reject, allowing for seamless integration of recommended tags and links into their knowledge cards. Continuous Learning: The recommendation system learns from user actions, improving the relevance of its suggestions over time based on user feedback and interaction patterns. User Control and Customization: Users have control over the recommendation settings, including the ability to prompt the system for suggestions at any time and customize the criteria used for generating recommendations.

Establish a confidentiality label so that cards that hold it are not displayed directly on the panel. This is useful for cases where I may want to have information that if I am working and my screen is visible to others, those cards are not visible

Currently, Recall automatically tags new content, but there is no way to retroactively update existing cards when new tags are created or when the tag structure changes. This leads to an inconsistent knowledge base where older cards miss out on newer, more relevant tags. I am proposing a "Tag Audit" feature (or a dedicated button) that would: Scan existing cards to apply newly created tags where relevant. Suggest global reorganizations to merge duplicates or fix "messy" hierarchies. Apply multiple relevant tags instead of just one, ensuring cards are correctly cross-referenced. This would ensure the knowledge base stays organized and up-to-date as the user's tagging system evolves over time.

Creating new tags erroneously due to use of upper or lower case for same-named tags: For example I have a tag named AI which is the base for many Sub Tags, example: I have Tag AI with Sub Tag /LLMS and I mistype it as Ai/LLMS, upon which the App creates a forked tag so now I have two root tags - AI and Ai - which forces me to have to fix the incorrect one to merge both AI/LLMS and Ai/LLMS. Suggestion, ask user if they one to create a new tag named Ai, or if they want to use the existing one.

Allow adding separator lines between groups of persona, e. G. Philosophy, psychology, sociology persona groups. This would add helpful groupings of different types of roles and expertise when determining the best approach.

Recall does this sometimes, but it is not consistent and in my experience does not happen most of the time. It would be nice if this happened consistently so that every video that is saved automatically has a "YouTube | [Channel Name]" as one of the tags.

...::: Summary Replace (or supplement) the current default view on launch with a configurable dashboard that gives users an at-a-glance overview of their knowledge base and actively resurfaces things they might otherwise never revisit. ...::: The problem Knowledge bases suffer from recency bias - users tend to interact only with whatever they added or searched for recently. Older content quietly becomes inaccessible, which defeats the purpose of building a second brain. A dashboard solves both the orientation problem (what's in here?) and the rediscovery problem (what have I forgotten?). ...::: Proposed default widgets (all toggleable, resizable) Knowledge base stats - total cards, tags, collections, and sources, with a sparkline showing growth over time. Recently added - the last 5–10 cards, with tag labels and a one-line preview. Recently visited tags - tags the user has opened in the last 7 or 30 days. Dormant tags - tags that haven't been opened in the longest time, to prompt rediscovery. Oldest cards - the first things ever saved, useful for seeing how thinking has evolved. Serendipity / Unexpected connections - see below. Today's random card - one card surfaced at random each day, a card-of-the-day feature. On this day - a list of cards added on this day over the years Knowledge gaps - tags or clusters with very few cards, suggesting where more research would be valuable. ...::: Serendipity / Unexpected connections widget - this includes everything in this FR: https://feedback.recall.it/feature-requests/p/serendipity-insights-on-load The concept: an AI-powered widget that periodically analyses the knowledge base and surfaces connections the user has never explicitly made - two cards or clusters that share themes, tensions, or implications that aren't obvious from the tags alone. Examples of what it might surface: "Your notes on DORA incident timelines and your notes on GDPR breach notification both deal with 72-hour windows - but they've never been linked" "Three cards tagged [AI regulation] and two tagged [employment law] share common ground around automated decision-making" This runs as a background process (daily or weekly, user-configurable) and lands results in the widget as a feed of insight cards the user can dismiss, save, or use to create explicit links. ...::: Customisation Drag-and-drop layout (grid system) Add/remove widgets from a widget library Widget-level refresh rate settings Light / dark / system theme for dashboard independent of the rest of the app Pinned searches or saved filters displayed as quick-launch tiles "Focus mode" - strips the dashboard to just two or three chosen widgets for distraction-free review sessions ...::: Additional ideas Weekly digest email generated from the dashboard - a Substack-style summary of what was added, what was dormant, and this week's serendipity finding Dashboard as homepage option vs. keeping it as a secondary view (both should be possible) Mobile-first widget designs - the dashboard should be genuinely useful on a phone, not just a desktop feature

Option to apply “AND” logic (instead of the current “OR”) when filtering by tags. This would allow users to see results that match all of the selected tags, rather than retrieving items that contain any selected tag. This offers more precise control when narrowing down results, ensuring that only tightly relevant content is shown.