How I Use Heptabase Journal (And When You Should Do It Differently)
The speaker demonstrates two approaches to using Heptabase's journaling feature: the common method of referencing cards from journal entries, and their preferred method of updating cards directly with dated notes. The direct-to-card approach provides better context and organization, though it requires improved search functionality to be fully effective.
Summary
The speaker begins by observing how most Heptabase users employ the journaling feature by creating dated bullet-point entries that reference specific cards, such as notes about a person. While this approach generates backlinks visible on the referenced card, it creates long, uncontextual lists of dates that require tedious searching through to find relevant information.
The speaker presents an alternative methodology: treating each subject card (like a person) as the single source of truth for that entity. Rather than creating journal entries that reference the card, they directly add timestamped notes within the card itself, using natural language date references like "today," "yesterday," or "last week." This keeps all information about a subject in one organized location.
A key advantage of this approach is bidirectional linking. When notes mention multiple entities (e.g., "I talked with person A about person B"), the speaker can navigate between cards and see the specific context of mention highlighted. The journal becomes a reference point for dates rather than the primary storage location. When viewing a date card, backlinks show all entities mentioned on that day, while viewing a person card shows all contextual information about that person.
The speaker acknowledges a significant limitation: Heptabase's search functionality doesn't adequately support searching within backlinks on the desktop version, making it difficult to filter results by type when many backlinks exist. They suggest improvements like collapsible search results organized by content type (whiteboards, notes, etc.) would enhance usability.
Key Insights
- The speaker observes that most users create dated bullet-point entries in the journal that reference other cards, which generates backlinks but creates long, uncontextual lists requiring tedious searching to find relevant information
- The speaker treats each subject card as the single source of truth and adds timestamped notes directly within those cards rather than creating separate journal entries, keeping all related information organized in one location
- When multiple entities are mentioned in a single dated note (e.g., interactions between person A and person B), Heptabase highlights the specific context of mention when navigating between related cards, enabling contextual understanding
- The speaker reports that Heptabase's search functionality on the desktop version does not search within backlinks when using the find function, unlike the browser version, making it difficult to locate specific information within extensive backlink lists
- The speaker suggests that organizing search results by content type with collapsible categories would help users focus on relevant information types, but this functionality does not currently exist in Heptabase
Topics
Transcript
[0:00] Over the time I saw many members how they use Heepbase and the journaling feature inside Heepbase and I thought it might be of interest for you how I'm using the journaling app as there are two ways to use it if you reference any other cards. Let's have a look here in my demo account on Hepabase and here we have the journal app and what I have seen is that people if they want to take notes about a specific person they just say person A this card and they make a bullet list like this. Okay, they take notes about this person on that day. [0:31] Then on another day, they do this again and they…
Full transcript available for MurmurCast members
Sign Up to AccessMore from ICOR with Tom | AI Productivity
Claude sucks. And here is why.
This video explores why Claude and other AI models produce inconsistent results, examining how model selection (Haiku vs Sonnet vs Opus vs Claude 3.5 Fable), effort levels, prompt clarity, and system context dramatically affect output quality. The speaker demonstrates these differences through a Venn diagram creation task and advocates for using organized folder systems with documented SOPs and code-based solutions to ensure consistent AI performance.
Claude can now prompt its own SESSIONS. I built the boss.
Claude now has the ability to interact with multiple sessions, allowing a streamlined orchestration of tasks through an agent called Larry. This integration can simplify task management and enhance productivity by enabling a single main session to oversee various sub-sessions or tasks.
I switched off Claude's memory. My folder does it better.
Tom explains why he disabled Claude's native memory feature and instead built a local folder-based system (myPKM) to maintain persistent context with AI agents. He shares an audit revealing systemic issues like contradictory documentation, forgotten rules, and parallel session conflicts, then outlines how to systematically fix these problems using the ICOR productivity methodology.
I do not download invoices anymore. My folder and Claude do it.
The speaker demonstrates how he uses AI (Claude) integrated with his local folder system to automate invoice collection for his accountant, downloading 36 invoices in 5 minutes with minimal manual intervention. He explains his AI team methodology, work streams, and how persistent local knowledge bases enable AI to handle recurring tasks independently without constant re-explanation.
Claude finished the work. Then I lost it in the chat.
A creator demonstrates a comprehensive local-folder-based system for managing AI-generated deliverables asynchronously, featuring an interactive browser interface that tracks decisions, changes, and conversations without relying on external databases or web applications. The system enables collaborative editing, visual annotations, and automatic context preservation across multiple parallel AI projects.