Card and Summary Content Merging
B
Brandon McIntosh
Sometimes more than one article or document may pertain to the same card (i.e., CrowdStrike articles are EVERYWHERE). I don't need 300 cards about the same topic, but each article has a little bit of detail that wasn't in the previous.
Solutions:
- Card Merging: Allow us the ability to create distinct cards for a given topic or tag, but then merge the summaries and tags into the same card after-the-fact.
This process could be performed by, and we should be given the option to either
a. appending one summary to the other, or; (makes content long, but retains original links and summaries)
b. merging summaries into some coherent, detailed summary of all linked documents (may lose some links and context)
- Content Merging: Allow us the ability to add additional links and documents to the same card and update the summaries based on the context from all given documents.
This process could be performed by, and we should be given the option to either:
a. Creating separate header sections for each reference pertaining to the same card (i.e. heading separation of distinct summaries of each resource)
b. Developing summaries based on the content of all supplied reference material.
- Content Extrapolation: The idea behind this software, at least for me, is that it allows me to see connections where I didn't originally see them. Rather than feeding everything into the same card, extract the content from the supplied reference, and
a. extrapolate and append summary sections into existing cards where it makes sense
b. create new cards for summary content where existing cards don't exist.
Log In
s
schroell
Additional Feature Request "Multi-Turn Chat Synthesis":
Extend the current single-response card creation functionality to support multi-turn chat consolidation. Currently, users can create a card from one chat response. The requested feature enables creating a consolidated card from an entire chat thread containing multiple questions and answers.
The system should use AI to analyze all questions and responses within a selected chat conversation, identify redundant and overlapping information from internal sources, and generate a comprehensive summary report. This consolidation would synthesize information scattered across multiple chat turns into a single, coherent overview of the explored topic.
The feature addresses deep-dive research scenarios where users iteratively refine questions and accumulate relevant information across several exchanges. Instead of manually combining fragmented responses, users would obtain an automatically generated, structured report that captures the complete context and findings from the entire conversation thread.
This enhancement transforms multi-turn AI chat sessions into actionable, consolidated knowledge artifacts based on the organization's internal sources.
L
Luay
Yes this would be huge. There are numerous multi-part videos on youtube where a single card makes sense to combine all the data.
m
monsalvo.geoffrey
Adding onto this videos and podcasts split in multiple parts as a use case for card merging