Engramma

Quickstart

beginner

Store and retrieve your first memory in under 5 minutes. Get your API key and start building.

Prerequisites

Step 1: Get your API key

  1. Sign in to app.engramma-memory.com
  2. Navigate to Settings → API Keys
  3. Click Create New Key
  4. Copy your key and store it securely
Warning

Never expose your API key in client-side code or commit it to version control. Use environment variables.

Step 2: Set your API key

export ENGRAMMA_API_KEY="your-api-key-here"

Step 3: Store your first memory

curl -X POST https://api.engramma-memory.com/v1/memory/text/store \
  -H "X-API-Key: $ENGRAMMA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "Paris is the capital of France"}'

Expected response:

{
  "success": true,
  "pattern_id": "pat_a7f2c1e9",
  "embedding_dim": 384,
  "patterns_used": 1,
  "patterns_limit": 5000
}

Step 4: Retrieve memories

curl -X POST https://api.engramma-memory.com/v1/memory/text/retrieve \
  -H "X-API-Key: $ENGRAMMA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "What is the capital of France?", "top_k": 5}'

Expected response:

{
  "results": [
    {
      "text": "Paris is the capital of France",
      "similarity": 0.94,
      "pattern_id": "pat_a7f2c1e9",
      "metadata": null
    }
  ],
  "latency_ms": 2.3
}

Step 5: Use intelligent recall

For higher-quality results that learn from usage patterns, use /v1/memory/text/recall:

curl -X POST https://api.engramma-memory.com/v1/memory/text/recall \
  -H "X-API-Key: $ENGRAMMA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "What is the capital of France?", "top_k": 5}'

Expected response:

{
  "results": [
    {
      "text": "Paris is the capital of France",
      "similarity": 0.96,
      "pattern_id": "pat_a7f2c1e9",
      "metadata": null
    }
  ],
  "info": {},
  "latency_ms": 8.1
}
Info

/text/recall uses Active Inference + semantic re-ranking and typically returns higher similarity scores than /text/retrieve. Results improve as the memory builds up access patterns over time.

What you just did

  1. Stored a piece of knowledge — the engine embedded it as a 384-dimensional vector
  2. Retrieved it using a natural-language query (semantic similarity, not exact match)
  3. Recalled it using the full cognitive engine (Active Inference)

What's next?

IntroductionHow It Works