It's Tuesday afternoon and the next patient on the schedule is somebody you haven't seen in nine months. The EHR shows the diagnosis codes from the last visit and a list of medications. What the EHR doesn't show is the conversation you had at the end of that visit about her brother moving in to help with caregiving, the worry she had about a new symptom that didn't fit any pattern, and the agreement you made to revisit the medication question once she'd talked to her cardiologist. You have ninety seconds before you walk in.
This post is about the layer of clinical context that lives outside the EHR — your own working notes, reference material, journal articles you've been meaning to come back to, and the running thread of detail that makes each visit a continuation rather than a restart. None of that belongs in the chart of record. All of it belongs somewhere you can actually find it.
A notes app is a place for a clinician's own structured, private working notes. Not a regulated platform, not a substitute for an EHR, not a record of clinical care for legal purposes — that's what the chart is for. This is the layer around it.
What clinicians need that an EHR doesn't give them
Every clinician we've talked to runs some version of the same parallel system on paper, in a notebook, in Apple Notes, in a shared drive. The reasons are consistent:
- The EHR is for documentation, not thinking. Clinical reasoning notes don't fit the templates.
- Reference material — guidelines, journal articles, dosing tables — lives in twelve browser tabs and a folder of saved PDFs that's never organized.
- Inter-visit continuity — the small contextual details that don't rise to the level of a chart note — has nowhere to go.
- Personal teaching notes, case reflections, and learning logs are useful career-long but die in inbox folders.
A notes workspace handles all of that, and an AI assistant inside that workspace makes it actually retrievable instead of accumulating into a different kind of mess. The same shape applies to specific disciplines — AI Notes for Dentists and AI Notes for Veterinarians follow the same pattern with different chart conventions.
Markdown pages, structured the way you think
In Docapybara, every page is plain markdown. That means each patient context page, each clinical reference, each journal-club summary is a normal text document — searchable, copyable, exportable.
A common shape clinicians use:
- A
Patients section with one nested page per active patient. Each page holds your working notes, the things-to-remember list, and the conversational context that doesn't fit a chart field.
- A
References section organized by condition, drug class, or specialty area — guideline summaries, dosing notes, decision-support logic written for yourself.
- A
Cases section for teaching notes, journal summaries, and case reflections you'd want to revisit during board prep or for a future presentation.
- An
Operations section for the practice-management side — staff schedules, supplier orders, the running list of things you've been meaning to fix in the workflow.
Page nesting goes as deep as needed. The structure evolves as the practice does, and because everything is plain markdown, restructuring is just a drag-and-drop operation that doesn't lose any content.
Letting the agent read across your reference library
The thing that changes the day-to-day experience is the assistant. Capy reads your vault on demand and uses 27 tools to do it. The kinds of questions that become answerable in seconds:
- "What does my reference page on heart-failure dosing say about starting a new patient on an SGLT2 inhibitor with eGFR around 35?" The agent finds the relevant page, locates the relevant section, and quotes it back.
- "Pull every patient I've seen in the last six months where I noted concern about medication adherence — give me a list with dates and what I wrote at the time."
- "Find every case note where I documented an unusual presentation that didn't fit the working diagnosis."
These aren't medical decisions — they're retrieval. The clinical thinking is still yours. What changes is how much of the working memory of the practice is actually accessible to you instead of buried under a year of accumulated notes.
Recording a consult for your own working notes
Some encounters benefit from an audio recording for your own follow-up — a long care-planning conversation, a complex patient handoff, a teaching session with a resident. Docapybara records audio in the workspace and transcribes with speaker labels, so you can see who said what.
The transcript is for your private notes — your own working memory, not the chart. When you go back to write a longer reflection or a teaching summary later, the conversation is captured as searchable text instead of relying on what you remember three days later.
For sensitive recordings, the usual professional considerations apply: get appropriate consent, follow your organization's policies, and treat your private notes the same way you'd treat any other private clinical material. The platform is a private workspace; the policy and process layer is yours.
PDFs of guidelines and articles, finally searchable
Most clinicians have a folder of PDFs — guidelines from professional societies, journal articles they've been meaning to read, dosing references, FDA labeling. They're useful when you can find them and frustrating when you can't.
Docapybara converts PDFs to markdown automatically when you drop them in. So when you ask "what does the latest hypertension guideline say about starting therapy in a patient over 75?", the agent searches across every PDF in your vault and pulls the relevant passage back as text. The source PDF is still one click away when you want to read the original.
This turns a passive archive into a working reference library. The articles you saved nine months ago and forgot become retrievable the moment they're relevant.
A live tracker for the things that follow you home
Every clinician has a running list of things-not-yet-done — patients to follow up with, results to chase, calls to return, learning items to revisit. That list is usually scribbled on the back of a printed schedule that gets shredded at the end of the week.
A :::database::: directive embeds a live database directly inside any markdown page. So your end-of-day page can include a follow-up tracker with columns for patient (or initials, depending on how you handle private notes), what's owed, due date, and status. Six column types are available, which covers most of what you'd want to track.
When you tell the assistant "add a follow-up for Mr. Chen — call about the labs by Thursday", it adds the row. When you ask "what's still open from last week?", it shows you. The list lives next to the prose context of the day, not in a separate to-do app you forget to open. For staff handoffs, Documenting Patient Interactions When You're Not the Clinician covers the front-desk side of the same workflow.
A reference structure that holds up over years
The hardest part of any clinician's notes system isn't the first month — it's year three. By then you have hundreds of patient pages, dozens of reference summaries, scores of case notes, and the structure that worked at the start has started to feel cramped.
Because Docapybara doesn't put a depth cap on page nesting, you can restructure as the practice grows without flattening. References → Cardiology → Heart Failure → Dosing is a normal path; so is References → Cardiology → Heart Failure → Cases → 2025 → Refractory if that's what you need. Every level is a real markdown page that can hold its own content, not a folder placeholder that's empty when you click in.
When the structure is stable, the agent's answers stay relevant. When you ask "summarize everything I have on the management of refractory heart failure in older adults", the agent walks the relevant subtree instead of scanning the whole vault. Faster answer, more focused context. The contrast with a folder-only tool is described in Docapybara vs Obsidian.
Try Docapybara free
To be clear about scope before the setup test: this is a private notes workspace for the layer of context that lives around clinical work. It is a place for your own working notes, your own reference material, your own thinking-out-loud about cases. It is not the medical record, and it is not a regulated platform handling protected health information at the level of an EHR or a documentation product. The boundary between what belongs in the chart and what belongs in your private notes is a clinical and regulatory judgment you make as the clinician — not something a notes app decides for you. What the workspace does well is what most clinicians actually struggle with: keeping the working layer organized, searchable, and useful three years from now. The chart of record handles the rest.
The setup test is small. Open Docapybara, create a page for the first patient on tomorrow's schedule, paste in whatever working notes you've got on them, and ask the agent for a three-line summary of what to remember from the last visit. Either it gives you something useful in fifteen seconds or it doesn't — and you'll know.
Try Docapybara free — bring your messiest reference folder, your collection of PDFs, and the running list you've been keeping on paper. See whether having all of it searchable in one place changes how you work.