v2.0.0 engram.courses

Synthesize academic research.
Own your memory models completely offline.

Engram is a private, multi-pass synthesis workspace and active recall engine designed for deep learning. No clouds, no telemetry, no tracking.

Detected Environment

Engram Studio for macOS

Download App Image (.dmg)
SHA-256 Verified Build View Source Vault
⚠️

macOS Security & Gatekeeper Guide

Because Engram is built locally and is not signed with a paid Apple Developer ID, macOS Gatekeeper will display a security warning when you open it for the first time. Depending on your system security policy, you will see one of the two warnings below:

Scenario 1: "Engram is damaged and can't be opened"

This is standard macOS behavior for unsigned binaries. To clear the quarantine flag and refresh the local signature, move Engram to your Applications folder, then run these two commands in your Terminal:

xattr -cr /Applications/Engram.app
codesign --force --deep --sign - /Applications/Engram.app
Scenario 2: "Developer cannot be verified"

If you get this standard warning, you can open the app without using the Terminal:

  1. Right-click (or Control-click) the Engram app icon inside your Applications folder and select Open.
  2. A confirmation prompt will appear. Click Open or Open Anyway.
01 / LOCAL SYNTHESIS MATRIX

Two-Pass Evaluation

Slices long academic files seamlessly into custom contextual layers before prompting specialized engines to aggregate dense structural ledgers.

02 / CORE ARCHITECTURE

Active Recall Loops

Automatically extracts high-density question and answer pairs structured cleanly into mathematical SM-2 interval scheduler engines.

03 / SECURE CONTAINERS

Complete Sovereign Files

Bypasses network sandboxing completely via native Tauri frameworks. Reads raw local directories, models, and databases strictly under your thumb.

Optimized Hardware Tiers

SYSTEM ALLOCATION RECOMMENDED LOCAL DECK TARGET FOOTPRINT
Apple Silicon 8GB (M1/M2) Phi-4 Mini 3.8B / Llama 3.2 3B ~2.3 GB VRAM
Unified Memory 16GB (M4 Pro) Gemma 4 12B Workstation Class ~7.6 GB VRAM
Workstation Tier 24GB+ Gemma 4 26B MoE (Mixture of Experts) ~16.0 GB VRAM