Quick Start Guide

From zero to structured audio intelligence in under 5 minutes.

Step 1 — Choose Your Plan

Plan	Price	Minutes/mo	Best for
Starter	$25/mo	1,000	Indie devs, testing
Professional	$99/mo	5,000	Teams, production apps
Dev Pro	$349/mo	25,000	Power users, solo builders
Enterprise	$1,595/mo	100,000	Teams of up to 5, business

Click Get started on your chosen tier — you'll be redirected to Stripe to complete payment. After payment, you'll receive an API key via email.

Step 2 — Get Your API Key

After subscribing, your API key is delivered to your inbox. It looks like this:

slo_abc123xyz789def456ghi ← your key, keep it private

Your key is also visible in the Stripe Dashboard under your subscription, or contact hello@signalloomai.com if you need it resent.

⚠️ Never share your API key publicly or commit it to version control. Use environment variables in production.

Step 3 — Make Your First Transcription

Option A — cURL (fastest)

      # Replace $SL_API_KEY with your key

      curl -X POST https://api.signalloomai.com/v1/transcribe \

        -H "Authorization: Bearer $SL_API_KEY" \

        -F "file=@meeting.mp4" \

        -F "speakers=true" \

        -F "entities=true"

Option B — Python SDK

      from signalloom import SignalLoom

      client = SignalLoom(api_key="slo_abc123xyz...")

      result = client.transcribe(

        file="./meeting.mp4",

        speakers=True,

        entities=True

      )

      print(result.knowledge)  # → structured JSON

Option C — JavaScript / Node.js

      import { SignalLoom } from '@signalloom/sdk';

      const client = new SignalLoom({ apiKey: 'slo_abc123xyz...' });

      const result = await client.transcribe({

        file: './meeting.mp4',

        speakers: true,

        entities: true

      });

      console.log(result.knowledge);  // → structured JSON

Step 4 — Understand the Output

Signal Loom doesn't just return plain text. You get a structured Knowledge Object — machine-readable JSON designed for AI consumption:

{
  "transcript": "Patient reports chest pain starting Tuesday...",
  "speakers": [
    { "id": "SPEAKER_00", "name": "Dr. Chen", "type": "physician" },
    { "id": "SPEAKER_01", "name": "Patient", "type": "patient" }
  ],
  "segments": [
    {
      "speaker": "SPEAKER_00",
      "start": 0.0,
      "end": 4.2,
      "text": "Tell me about the chest pain."
    }
  ],
  "entities": [
    { "type": "SYMPTOM", "value": "chest pain", "speaker": "SPEAKER_01", "start": 2.1 },
    { "type": "DATE", "value": "Tuesday", "speaker": "SPEAKER_01", "start": 5.3 }
  ],
  "topics": [
    { "label": "Chief Complaint", "start": 0.0, "end": 45.2 }
  ]
}

This structure plugs directly into RAG pipelines, agent memory, knowledge graphs, or any downstream AI system — no parsing required.

Step 5 — API Reference

The Signal Loom API is REST-based and lives at:

https://api.signalloomai.com/v1/

Endpoints

POST /v1/transcribe — Upload media, get full pipeline (transcription + speakers + entities)
GET /v1/health — Service health check
GET /v1/stripe/prices — View available pricing tiers
POST /v1/stripe/checkout — Create Stripe checkout session

Request Options

Parameter	Type	Default	Description
file	multipart	—	Audio or video file (mp3, mp4, wav, m4a, flac)
speakers	bool	false	Enable speaker diarization
entities	bool	false	Enable named entity extraction
topics	bool	false	Enable topic segmentation

Step 6 — Supported File Formats

Signal Loom accepts the following formats natively:

Type	Formats
Audio	mp3, wav, m4a, flac, ogg, webm, aac
Video	mp4, mov, avi, mkv, webm

💡 No file size limit enforced at the API layer — limits depend on your plan's concurrent job allowance. For files over 500MB, contact hello@signalloomai.com.

Troubleshooting

Getting a 401 Unauthorized

Your API key is missing, expired, or invalid. Check that your Authorization: Bearer header matches exactly — no extra spaces or newlines.

Getting a 429 Rate Limited

You've hit your plan's request-per-minute or concurrent-job limit. Upgrade your plan or wait before retrying.

Transcription is slow

Longer files take proportionally longer. A 30-minute file typically completes in 60–90 seconds. For real-time needs, consider our streaming mode (coming soon).

Speaker labels are wrong

Diarization identifies number of speakers, not their names. Named speakers require post-processing using the speaker labels with your own identification logic.

File won't upload

Check the file format is in the supported list above. Also check the file isn't corrupted — try playing it locally first.

Need Help?

Email us anytime at hello@signalloomai.com — we respond within one business day.

For billing and subscription questions, your receipt and subscription details are available in the Stripe Dashboard.