Connect all your tools to Andebet.

Ingest call recordings from Twilio, stream audio from S3 buckets, expose transcripts to Claude via MCP, or build custom workflows with our official Python and TypeScript SDKs.

12+
Ready Connectors
Cloud, telephony, AI & chat
HMAC-SHA256
Signed Webhooks
Secure asynchronous callbacks
MCP 1.0
Agent Native
Claude Desktop & Cursor support
Typed SDKs
Python & TypeScript
Auto-retry & stream support

Three ways to connect your stack.

Choose the architectural layer that best fits your engineering workflow.

Event-Driven Webhooks

Never poll endpoints. Get instant callbacks with HMAC signatures when audio is finished transcribing or when speaker segmentation is complete.

High-Throughput REST API

Standard OpenAPI 3.1 endpoints. Upload audio files up to 2GB, configure dialect hints, and receive timestamped Ge'ez and Latin text.

Model Context Protocol

Turn your audio transcripts into queryable tool functions for Claude Desktop, Cursor, and autonomous agent loops.

Production connectors and developer libraries.

Choose from native protocols, cloud storage hooks, telephony pipelines, and programming SDKs.

Official ProtocolAI & Agents

Model Context Protocol (MCP)

Expose your audio transcripts, speaker turns, and corpus search directly to Claude Desktop, Cursor, and any MCP-compliant agent framework.

Claude DesktopCursor IDEnpx @andebet/mcp-server
Call CenterTelephony

Twilio Voice & SIP

Stream live dual-channel calls or batch-process recorded MP3/WAV voicemails for automated customer QA and Amharic sentiment analysis.

Streaming MediaCall Recording WebhooksDual-Channel
Object StorageCloud Storage

Amazon S3

Automatically trigger transcription jobs whenever audio or video files are deposited in an S3 bucket via SQS or EventBridge notifications.

Auto-IngestionIAM Role AuthPresigned URLs
Voice NotesMessaging

WhatsApp Business API

Transcribe customer voice messages in Amharic, Afaan Oromo, and Tigrinya into instant text replies for chatbots and support agents.

Opus / OGG AudioMeta Cloud APIInstant Webhooks
Voice MemosMessaging

Telegram Bot API

Build Telegram bots that convert incoming OGG voice memos into Ge'ez and Latin text, summaries, and sentiment classifications.

OGG VoiceBot Token AuthWebhook Routing
GCP IntegrationCloud Storage

Google Cloud Storage

Stream high-volume broadcast audio from GCS buckets with native Pub/Sub push webhooks and service account token authorization.

Pub/Sub PushGCP BucketsDirect Streaming
pip install andebetSDKs & APIs

Python SDK

Full Python bindings with type hinting, streaming generator support, chunked multi-part upload helpers, and automatic retries.

pip install andebetAsyncIOPydantic Models
npm i andebetSDKs & APIs

TypeScript / Node SDK

Universal TypeScript client for Next.js, Node.js, Bun, and browser environments. Includes HMAC webhook verification utilities.

npm i andebetEdge CompatibleTypeScript Types
HMAC-SHA256SDKs & APIs

Async Signed Webhooks

Receive real-time notifications for transcription.completed, speaker.diarized, and analysis.ready events with replay-attack protection.

HMAC ValidationExponential BackoffEvent Logs
VoIP PBXTelephony

Asterisk & FreePBX

Integrate on-premise PBX recording queues with Andebet via local curl scripts or our lightweight Docker proxy daemon.

AGI ScriptsWAV Mono/StereoPrivate VPC
Meeting SyncMessaging

Slack Workflow Integration

Share automated meeting minutes, Ge'ez action item lists, and speaker summaries directly to team Slack channels.

Incoming WebhooksInteractive BlocksDaily Briefings
Postgres + AudioCloud Storage

Supabase Storage & DB

Store raw audio in Supabase Storage and write timestamped Ge'ez transcripts back to PostgreSQL tables using Edge Functions.

Edge FunctionsStorage Webhookspgvector Search
webhook_signature_validator.py
import hmac
import hashlib
from flask import Flask, request, jsonify

app = Flask(__name__)
WEBHOOK_SECRET = "whsec_andebet_demo_key_98765"

@app.route("/webhooks/andebet", methods=["POST"])
def handle_andebet_webhook():
    signature = request.headers.get("X-Andebet-Signature")
    timestamp = request.headers.get("X-Andebet-Timestamp")
    
    # Verify HMAC SHA-256 signature
    payload = f"{timestamp}.{request.get_data(as_text=True)}"
    expected = hmac.new(
        WEBHOOK_SECRET.encode(),
        payload.encode(),
        hashlib.sha256
    ).hexdigest()
    
    if not hmac.compare_digest(signature, expected):
        return jsonify({"error": "Invalid signature"}), 401

    data = request.json
    print(f"Transcription finished: {data['job_id']}")
    print(f"Language: {data['detected_language']} | Duration: {data['duration_seconds']}s")
    
    return jsonify({"status": "acknowledged"}), 200

Frequently Asked Questions

Common questions on connecting Andebet to your applications, storage, and developer workflows.

When an audio processing job finishes, Andebet sends an HTTP POST request to your specified endpoint. Each payload is signed with an HMAC-SHA256 signature in the X-Andebet-Signature header to ensure authenticity and protect against replay attacks.

Yes. You can provide presigned URLs or configure AWS IAM Role delegation / GCP service account authorization so Andebet streams audio directly from your private bucket without public internet exposure.

Our native MCP server runs locally via npx @andebet/mcp-server or in a Docker container. Claude Desktop and Cursor can directly search your transcripts, inspect speaker turn ratios, and run semantic queries across your Ethiopian audio corpus.

Yes. We maintain official SDKs for Python (pip install andebet) and TypeScript / Node.js (npm i andebet) with full TypeScript types, streaming generators, and built-in retry logic.

Connect Ethiopian speech intelligence in minutes.

Sign up for free developer access and receive 60 complimentary minutes with instant API keys and SDK support.

  • 45 free minutes included
  • No credit card required
  • Instant API & Workspace access
  • Enterprise data sovereignty