Private Mac notes for academics & researchers

Academics, doctoral researchers, postdoctoral fellows, and principal investigators operate at the frontier of human knowledge discovery. In any given academic term, a researcher synthesizes dense peer-reviewed literature, participates in departmental seminars and international colloquiums, supervises graduate research students, and drafts competitive grant proposals.

The lifeblood of scholarship is the cumulative synthesis of ideas over time: connecting an experimental observation in a laboratory meeting with a theoretical framework published a decade earlier, or tracing a methodological debate across dozens of journal articles.

Yet mainstream productivity tools actively sabotage academic research. Commercial cloud note-taking platforms trap literature notes inside proprietary web databases, mangle complex mathematical equations, and lock your intellectual life’s work behind recurring subscription paywalls. Furthermore, using commercial AI meeting recorders in academic seminars introduces awkward cloud bots that distract speakers and upload unreleased scientific hypotheses to third-party corporate servers.

Oyma is the dedicated academic workstation engineered for scholars on Apple Silicon Macs. Operating locally on your hardware, Oyma captures academic seminars silently via ScreenCaptureKit, runs speech recognition 100% on-device with local Whisper models, and organizes literature reviews and thesis chapters in an interconnected Markdown knowledge graph.

The academic research barrier: Why cloud note silos fail scholars

For scholars engaged in long-term intellectual work, relying on generic SaaS productivity software introduces severe operational and epistemological friction:

1. The proprietary archive trap: Academic research spans decades. Storing dissertation notes or literature syntheses in cloud proprietary apps (like Notion or Evernote) risks catastrophic data loss if the vendor changes pricing, goes bankrupt, or alters file formats. Scholars require future-proof plain Markdown on disk.
2. Bot disruption in academic colloquiums: When an automated bot enters an intimate research symposium or departmental defense announcing that a cloud AI is recording, it chills academic debate and violates seminar etiquette.
3. Premature disclosure of scientific hypotheses: Academic lab meetings frequently discuss unreleased experimental findings, grant proposal strategies, and patentable discoveries. Transmitting this raw audio to commercial cloud transcription vendors risks intellectual property leakage.
4. Weak scientific synthesis tools: Most note apps lack native bidirectional linking, making it impossible to construct an emergent, interconnected literature graph where concepts link across authors, years, and methodologies.

The academic researcher’s literature and seminar topology

Oyma organizes your scholarship into an interconnected, sovereign research graph that links literature reviews, experimental findings, and seminar discussions:

┌────────────────────────────────────────────────────────────────────────┐
│                   ACADEMIC RESEARCH KNOWLEDGE TOPOLOGY                 │
├────────────────────────────────────────────────────────────────────────┤
│                                                                        │
│                    [[Thesis: Neural Representation Drift]]             │
│                    dissertation/chapter-03-drift-mechanisms.md         │
│                                ▲                                       │
│          ┌─────────────────────┼─────────────────────┐                 │
│          │                     │                     │                 │
│          ▼                     ▼                     ▼                 │
│  [Paper: Miller et al. 2024] [Lab Meeting: Oct 2026] [Symposium Keynote]│
│  Continuous manifold drift   Drift in motor cortex   Mathematical model│
│  `[@miller2024cortical]`     Empirical recording data Non-linear dynamics│
│          │                     │                     │                 │
│          └─────────────────────┼─────────────────────┘                 │
│                                ▼                                       │
│                    [[Concept: Manifold Stabilization]]                 │
│                    concepts/manifold-stabilization.md                  │
│                                                                        │
└────────────────────────────────────────────────────────────────────────┘

Using bidirectional [[wikilinks]], each seminar transcription and paper summary links back to core theoretical concepts. When drafting thesis chapters or peer-reviewed journal submissions, you can traverse the graph view to discover latent connections between seemingly unrelated experimental observations.

The 4-step scholarly workflow

1. Invisible seminar and colloquium capture

When you attend a guest lecture, departmental defense, or collaborative lab meeting on Zoom, Google Meet, or Microsoft Teams, Oyma identifies the active audio stream. One click activates ScreenCaptureKit capture. No bot enters the call, ensuring total academic decorum.

2. Live literature and seminar notation

While listening to the lecture, you can type live notes, insert LaTeX formulas ($$\mathcal{L}_{\text{drift}}$$), and record citekey references directly in the Live Preview editor.

3. On-device Whisper transcription across 13 languages

Following the symposium, Oyma‘s on-device Whisper engine transcribes the recording directly on your Mac’s Apple Silicon Neural Engine. The model accurately processes technical scientific terminology and multilingual discussions with zero cloud API costs.

4. Permanent vault filing in open Markdown

The transcribed seminar note writes directly to your local academic vault as a standard .md file. You can link key claims to your literature notes using [[wikilinks]] and citekey references.

Authentic academic seminar and literature note sample

Here is a realistic academic research note produced in Oyma:

---
title: "Lab Seminar: Manifold Geometry & Neural Population Dynamics"
date: 2026-10-08
type: seminar-notes
speaker: "Prof. Julian Vance (Computational Neuroscience Lab)"
citekey: "[@vance2026manifold]"
tags: [neuroscience, computational, manifolds, motor-cortex, seminar]
---

## Seminar Abstract & Theoretical Premise
Prof. Vance presented recent empirical findings on neural population manifold geometry during motor skill acquisition. Challenged the static manifold hypothesis by demonstrating low-dimensional drift that preserves task-relevant subspace projections over 6-month recording intervals.

## Mathematical Formulation & Dynamics
The population firing rate vector $\\mathbf{r}(t) \\in \\mathbb{R}^N$ is projected onto a low-dimensional manifold subspace via projection matrix $\\mathbf{W} \\in \\mathbb{R}^{d \\times N}$:

$$\\mathbf{z}(t) = \\mathbf{W} \\mathbf{r}(t) + \\mathbf{b}$$

Where task decoding accuracy remains invariant under rotation $\\mathbf{R} \\in SO(d)$ such that:

$$\\mathbf{z}'(t) = \\mathbf{R} \\mathbf{z}(t)$$

## Key Seminar Discussion Points & Critiques
- **Electrode drift vs representational drift:** Lab members raised concerns regarding chronic recording instability. Prof. Vance addressed this by showing identical drift dynamics across calcium imaging and Neuropixels probes.
- **Connection to dissertation:** This directly supports my hypothesis in [[Thesis: Neural Representation Drift]] regarding compensatory synaptic plasticity in downstream premotor circuits.

## Literature Backlinks & Citations
- Relates to: `[@kaufman2014cortical]` on preparatory activity subspaces
- Contradicts: `[@gallego2017neural]` static manifold preservation assumptions

## Action Items & Research Follow-up
- [ ] Researcher: Re-run PCA subspace alignment script on our lab's monkey dataset
- [ ] Researcher: Request pre-print PDF of Vance lab's upcoming Nature Neuroscience paper
- [ ] Researcher: Incorporate rotation invariance proof into Chapter 3 draft

Absolute data sovereignty for unreleased research

Scholars frequently handle vulnerable preliminary findings: unreleased experimental results, confidential peer review comments, and competitive grant proposals. Trusting this intellectual capital to cloud note apps exposes groundbreaking work to data leaks and vendor obsolescence.

Oyma guarantees total intellectual sovereignty:

• Zero cloud egress: Audio recording and Whisper transcription execute 100% on-device on Apple Silicon.
• Auditable security: Inspect the transparent Egress Log at any time to verify zero external network calls.
• Decades-long durability: Notes live as open .md files on your local drive, fully accessible in Obsidian, Zotero, LaTeX editors, and terminal tools forever.

Elevate your scholarship with rigorous literature synthesis, effortless seminar transcription, and sovereign Markdown notes.