Quickstart

Prerequisites

You’ll need an Engramme API key with access to the text source type. For the simplest setup, create an All Sources key in app.engramme.com.

export ENGRAMME_API_KEY="YOUR_API_KEY"

Step 1: Check API Health

First, verify the API is reachable by checking the health endpoint:

curl https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/health

You should see:

{
  "status": "healthy",
  "message": "Memory Machines Platform API is running."
}

Step 2: Submit Your First Document for Training

Create a sample text file:

cat > meeting-notes.txt <<'EOF'
Meeting Notes - January 15, 2025

Attendees: Sarah, Mike, Jennifer

We discussed the Q1 roadmap today. Sarah presented the new authentication
feature timeline - targeting end of February for beta. Mike raised concerns
about the current API response times, suggesting we prioritize performance
optimization before the launch.

Action items:
- Sarah: Finalize auth feature specs by Friday
- Mike: Run performance benchmarks this week
- Jennifer: Schedule follow-up for next Tuesday
EOF

Now submit it to the LMM training pipeline:

curl -X POST https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/memorize \
  -H "x-api-key: $ENGRAMME_API_KEY" \
  -F "file=@meeting-notes.txt" \
  -F "user_name=Your Name" \
  -F "source_type=text"

import os
import requests

url = "https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/memorize"
headers = {"x-api-key": os.environ["ENGRAMME_API_KEY"]}

with open("meeting-notes.txt", "rb") as f:
    files = {"file": f}
    data = {"user_name": "Your Name", "source_type": "text"}
    response = requests.post(url, headers=headers, files=files, data=data)

print(response.json())

const apiKey = process.env.ENGRAMME_API_KEY;
const formData = new FormData();
formData.append('file', new Blob(['your content'], { type: 'text/plain' }), 'notes.txt');
formData.append('user_name', 'Your Name');
formData.append('source_type', 'text');

const response = await fetch(
  'https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/memorize',
  {
    method: 'POST',
    headers: { 'x-api-key': apiKey },
    body: formData
  }
);

console.log(await response.json());

Choose a source type

The source_type field tells Engramme where the content came from. For this guide, use text. See Source Types for the full list.

API keys can be scoped to specific source types. If your key does not include text, this training request will be rejected.

You’ll receive a response like:

{
  "status": "success",
  "user_id": "your_user_id",
  "item_id": "abc123",
  "workflow_execution": "projects/.../executions/exec-xyz"
}

/v1/memorize returns when asynchronous training has started. Memories may take a short time to become recallable.

Step 3: Recall Memories

Wait briefly for training to run, then retrieve relevant memories:

curl -X POST https://memorymachines-gateway-prod-btf57kda.uc.gateway.dev/v1/recall \
  -H "x-api-key: $ENGRAMME_API_KEY" \
  -F "text=meeting"

Response:

{
  "memories": [
    {
      "custom_id": "abc123_memory_0",
      "item_id": "abc123",
      "score": 0.94,
      "content": {
        "narrative": "I attended a meeting with Sarah, Mike, and Jennifer to discuss the Q1 roadmap. Sarah presented the authentication feature timeline targeting end of February. Mike raised concerns about API response times.",
        "participants": ["Sarah", "Mike", "Jennifer"],
        "when": "January 15, 2025",
        "where": null,
        "tags": ["meeting", "Q1", "roadmap", "authentication"]
      }
    }
  ],
  "qa": [],
  "chunks": []
}

If no memories are returned immediately, retry after a short delay. Very short or synthetic training inputs may not always produce a recallable memory.

Next Steps

Authentication

Learn about API key management

Source Types

Pick the right source_type

Common Issues

401 Unauthorized

Your API key is invalid or missing. Make sure you’re including the x-api-key header with your request.

No memories returned

Training takes time. Wait a moment and try again. You can also check whether the document was accepted by looking at the response from /v1/memorize.

400 Bad Request

Check that your file and parameters are valid. For text training inputs, files must be UTF-8 encoded and under 10MB. Also make sure source_type is one of the supported values listed above.

Get Started

Console Setup

API Reference

SDKs

Prerequisites

Step 1: Check API Health

Step 2: Submit Your First Document for Training

Choose a source type

Step 3: Recall Memories

Next Steps

Authentication

Source Types

Common Issues

​Prerequisites

​Step 1: Check API Health

​Step 2: Submit Your First Document for Training

​Choose a source type

​Step 3: Recall Memories

​Next Steps

Authentication

Source Types

​Common Issues

Prerequisites

Step 1: Check API Health

Step 2: Submit Your First Document for Training

Choose a source type

Step 3: Recall Memories

Next Steps

Common Issues