Skill UI

This guide provides minimal working examples for Deepgram speech-to-text transcription. It demonstrates how to transcribe audio from a URL or a local file using both TypeScript and Python. Users can learn basic API patterns, implement advanced features like diarization (speaker identification), and compare various model options (e.g., nova-3, nova-2). Ideal for initial integration and testing.

A comprehensive runbook for standardizing incident response to Deepgram production issues. It guides users through automated triage (checking status, API connectivity, and model performance), severity classification (SEV1-SEV4), and implementing mitigation strategies, including building fallback services and replaying failed requests to ensure business continuity during outages or degradation.

A comprehensive guide for installing and configuring the Deepgram SDK for both Node.js and Python environments. This skill covers best practices for secure API key management using environment variables, client initialization (including version differences), and connection verification. Use this when integrating Deepgram's advanced speech-to-text and text-to-speech capabilities into a new application.

This guide establishes a robust local development environment for Deepgram integrations. It covers setting up test fixtures using sample audio, mocking API responses for isolated unit testing, configuring watch-mode development servers, and running comprehensive unit and integration tests using Vitest. Use this setup when initiating a Deepgram feature or requiring a reliable, fast development cycle.

This guide offers a comprehensive deep dive into migrating existing voice transcription workflows from multiple major providers, including AWS Transcribe, Google Cloud STT, OpenAI Whisper, and others, to Deepgram. It details feature mapping, best practices, and implements an adapter pattern, enabling seamless, reliable, and validated service switching.

This skill provides a comprehensive and isolated setup for managing Deepgram API integrations across different deployment stages (Development, Staging, Production). It includes typed configuration management, a client factory pattern, and Docker Compose profiles, ensuring that each environment uses specific credentials, models, and feature settings (like diarization or PII redaction) without conflict. Ideal for robust, multi-stage application deployment.

Implements a comprehensive observability stack for Deepgram integrations. It utilizes Prometheus metrics for tracking request rates, latency, and costs, OpenTelemetry for distributed tracing, Pino for structured logging, and AlertManager rules for proactive alerting. Essential for production monitoring, dashboard setup, and performance analysis of Deepgram API usage.

This guide provides advanced techniques to optimize Deepgram API performance. Learn how to preprocess audio using ffmpeg (resampling, noise reduction), select optimal transcription models (speed vs. accuracy), and implement efficient streaming or parallel batch processing to achieve high throughput and low latency.

A comprehensive checklist and guide for deploying Deepgram integrations into a production environment. It covers critical areas including robust authentication practices, system resilience design (e.g., backoff strategies, circuit breakers), performance optimization (singleton clients, audio preprocessing), security measures (PII redaction), and mandatory monitoring setup (Prometheus metrics and alert rules). Essential for any go-live phase.

This utility provides comprehensive patterns for reliably integrating with the Deepgram API. It implements concurrency-aware rate limiting (using p-limit), exponential backoff with jitter for retries, and a circuit breaker pattern. Use this when building high-volume, mission-critical applications that must handle API quotas, 429 rate limit errors, and temporary service disruptions gracefully.

Provides reference architectures for building high-scale, reliable transcription systems using Deepgram. It covers four main patterns: synchronous REST for short files, async queues (BullMQ) for batch processing, WebSocket for real-time streaming, and a hybrid router for mixed workloads. Use this when designing production-grade audio processing pipelines.

Provides production-ready coding patterns for integrating Deepgram using TypeScript and Python. This guide covers essential best practices, including implementing singleton client management, performing text-to-speech synthesis with Aura, and setting up advanced audio intelligence pipelines (e.g., summarization, topic detection, sentiment analysis). Use this when establishing robust Deepgram integration standards, refactoring existing SDK usage, or migrating to newer SDK versions.