Get AI summaries of any video or article — Sign up free
One Year Later with Nicole van der Hoeven - Session 2 at the Sketch Your Mind Conference, 2025 thumbnail

One Year Later with Nicole van der Hoeven - Session 2 at the Sketch Your Mind Conference, 2025

5 min read

Based on Zsolt's Visual Personal Knowledge Management's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Nicole van der Hoeven’s hybrid-note template makes every new note include both text and Excalidraw drawing space, removing the need to decide whether a note “should” be visual.

Briefing

Nicole van der Hoeven’s year-long shift in personal knowledge management centers on one practical decision: make every new note a “hybrid” by default—text plus Excalidraw drawing space—so switching between writing and sketching never becomes a separate workflow. After converting her approach, she reports that the friction disappears so completely that she now creates notes that are purely visual more often than expected, while still keeping text searchable inside the drawing.

The core mechanics start with templates inside Obsidian. Using the Templater plugin and Excalidraw integration, she sets up a new-note template that automatically provides both a markdown/text side and an Excalidraw side. A hotkey flips between modes, and the template reduces the mental question of whether a note “deserves” visuals—every note already has room for them. She also keeps multiple hybrid templates (for example, for people) with front matter and Excalidraw-related metadata, but the principle stays the same: hybrid is the default state.

From there, her workflow splits into two directions. First is “going backwards” by treating visuals as modular components. She builds libraries—especially an icon library—by scanning her vault for assets labeled with a naming pattern (e.g., “icon-”). Those icons function like notes: when an icon appears inside a larger visual, Obsidian backlinks let her trace where each component is used. She gives a telemetry-collector example where a “pipeline” component wasn’t a standalone note, yet it still behaves like one in her system because it’s embedded as a visual constituent and connected through backlinks. This encourages iterative refinement: if she later realizes a component should have been reused elsewhere, she can find it through the library.

Second is “going forwards,” assembling smaller pieces into larger diagrams. She describes embedding a telemetry-collector visual into an “observability” page, so the larger concept inherits traceability back to the smaller parts—links work in both directions, not just top-down. She also deconstructs existing architecture diagrams (from Loki marketing materials) into linked subpages, turning a single slide into a navigable system of concepts.

She addresses practical constraints that come with visual PKM at scale. Excalidraw notes don’t noticeably slow local vault performance, but they can slow publishing and can hit Obsidian Publish storage quotas. She also notes that image formats matter: she prefers SVGs when possible, uses scripts like “Shade Master” to recolor SVGs, and avoids backing up images to GitHub because they’re “hopelessly bloated,” relying instead on other sync/backups.

Beyond diagrams, she applies the same hybrid logic to conferences (embedding PDFs, screenshots, and photos into Excalidraw-based conference documentation), reference notes (book notes with short-form summaries plus per-chapter visuals), and even “observability for her own PKM.” That last experiment pipes note metadata through a Python/cron-to-JSON-to-Alloy-to-Loki-to-Grafana stack to track writing activity over time.

The session’s takeaway is less about drawing talent and more about system design: if hybrid notes are effortless to create and components are traceable through backlinks, visual thinking scales from individual ideas into maintainable knowledge structures.

Cornell Notes

Nicole van der Hoeven’s PKM workflow makes every new note hybrid by default: Obsidian templates automatically create both a text/markdown side and an Excalidraw drawing side. Hotkeys let her flip modes instantly, eliminating the “should this be visual?” decision and enabling notes that are sometimes purely visual while remaining searchable because Excalidraw text is plain text. She treats visuals as modular building blocks: an icon library is generated by scanning the vault for labeled assets, and backlinks preserve traceability from components to the larger diagrams that use them. She then scales up by embedding smaller visuals into bigger concept pages (e.g., telemetry collectors inside observability), creating two-way navigation between parts and systems. The approach matters because it turns visual thinking into a maintainable, linkable knowledge system rather than isolated sketches.

Why does making every note hybrid reduce friction so much, and what’s the concrete mechanism behind it?

The friction drops because the workflow removes the decision point. Instead of choosing between “text note” or “drawing note,” the template always provides both. In Obsidian, she uses the Templater plugin plus Excalidraw integration so a new note is born with markdown/text and an Excalidraw canvas. A hotkey toggles between modes (markdown ↔ Excalidraw), so switching is muscle memory. She also notes that the Excalidraw image side isn’t fully created until she switches to it, which avoids unnecessary work until it’s needed.

How do visuals become “modular components” that behave like notes rather than static drawings?

She relies on backlinks and vault indexing. Her icon library is generated by a script that scans the vault for files labeled with a pattern like “icon-…”, then surfaces those icons as reusable components. When an icon is used inside a larger Excalidraw note, Obsidian’s backlinks let her see where that component appears. That means she can “go backwards” from a concept (like a pipeline icon) to the larger visual(s) that used it, even if she never created a standalone note for that component.

What does “going forwards” look like in her system, and how does traceability survive the assembly?

Going forwards means embedding smaller visuals into larger concept pages. For example, she embeds a telemetry-collector visual into an “observability” page. Clicking through shows the embedded visual is linked in both directions (observability ↔ telemetry collector), and backlinks also capture other mentions (like a meeting note or an article reference). The result is that larger diagrams don’t become dead ends; they remain navigable maps of linked sub-concepts.

What scaling problems show up when using lots of Excalidraw notes, and how does she mitigate them?

Local vault performance is mostly fine, but publishing can slow down and storage quotas can be hit on Obsidian Publish. She also manages backup strategy: she backs up text to GitHub but avoids backing up images because they’re “hopelessly bloated.” For images, she uses other sync/backups such as Dropbox and Obsidian Sync, effectively keeping multiple copies while keeping GitHub lean.

How does she handle design consistency for icons across many diagrams?

She prefers SVG icons because they can be recolored and styled. She keeps black and red versions when using Flat icon assets, but later she adopts the Shade Master script in Excalidraw to skin a single SVG as needed while preserving the underlying links to that image asset.

How does she extend hybrid visual PKM beyond diagrams into tracking and documentation?

She documents conferences by embedding PDFs, screenshots, and photos into Excalidraw-based conference pages, creating a visual record of the event experience. For personal analytics, she treats her PKM like a monitored system: a Python script (cron every ~5 minutes) parses new/modified Obsidian files into JSON logs, Alloy tails the logs, sends them to Loki, and Grafana visualizes the results. That dashboard tracks metrics like recently updated notes and could be extended to links-per-note trends.

Review Questions

  1. What specific template and hotkey setup makes switching between text and Excalidraw “frictionless,” and why does that matter for habit formation?
  2. How do backlinks and vault indexing turn icons/visual components into reusable building blocks that support both bottom-up and top-down knowledge building?
  3. What trade-offs appear when scaling visual notes (especially for publishing), and what backup/sync choices does she make to manage them?

Key Points

  1. 1

    Nicole van der Hoeven’s hybrid-note template makes every new note include both text and Excalidraw drawing space, removing the need to decide whether a note “should” be visual.

  2. 2

    Hotkey-based mode switching (markdown ↔ Excalidraw) is central to keeping the workflow fast enough to sustain habit.

  3. 3

    Visual components become reusable “building blocks” when backlinks preserve traceability from embedded parts (like icons) to the larger diagrams that use them.

  4. 4

    Complex systems scale by embedding smaller visuals into larger concept pages, keeping two-way navigation between parts and whole ideas.

  5. 5

    She prefers SVGs for icons because they can be recolored/styled (e.g., via Shade Master), improving consistency without breaking asset links.

  6. 6

    Excalidraw-heavy vaults may slow publishing and can hit Obsidian Publish storage quotas, so she plans around publishing limits.

  7. 7

    She applies the same hybrid approach to conferences, reference notes, and even PKM observability dashboards built with Python, Loki, and Grafana.

Highlights

Hybrid-by-default templates eliminate the “text or drawing?” decision—she reports creating purely visual notes more often than expected.
An icon library generated from vault assets plus backlinks lets visual components behave like navigable notes, even when they aren’t standalone pages.
Embedding a telemetry-collector visual inside an observability page preserves traceability through backlinks, creating two-way conceptual navigation.
Publishing can become the bottleneck (storage quotas and slower publishing), while local vault performance stays mostly responsive.
Her PKM observability experiment treats note metadata like system telemetry, piping it into Loki and visualizing it in Grafana.

Topics

Mentioned