deepgram-core-workflow-a
Implement production pre-recorded speech-to-text with Deepgram. Use when building audio transcription, batch processing, or implementing diarization and intelligence features. Trigger: "deepgram transcription", "speech to text", "transcribe audio", "batch transcription", "deepgram nova", "diarize audio".
- License 许可
- MIT
- 最近更新
- 2026-07-13
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- deepgram-core-workflow-a
- tags
- [ "saas", "deepgram", "voice-ai", "transcription", "workflow", "stt" ]
- author
- Jeremy Longshore <jeremy@intentsolutions.io>
- license
- MIT
- version
- 1.0.0
- description
- Implement production pre-recorded speech-to-text with Deepgram. Use when building audio transcription, batch processing, or implementing diarization and intelligence features. Trigger: "deepgram transcription", "speech to text", "transcribe audio", "batch transcription", "deepgram nova", "diarize audio".
- allowed-tools
- Read, Write, Edit, Bash(npm:*), Bash(pip:*), Grep
- compatible-with
- claude-code, codex, openclaw
03 SKILL.md 正文
Deepgram Core Workflow A: Pre-recorded Transcription
Overview
Production pre-recorded transcription service using Deepgram's REST API. Covers transcribeUrl and transcribeFile, speaker diarization, audio intelligence (summarization, topic detection, sentiment, intent), batch processing with concurrency control, and callback-based async transcription for large files.
Prerequisites
@deepgram/sdkinstalled,DEEPGRAM_API_KEYconfigured- Audio files: WAV, MP3, FLAC, OGG, M4A, or WebM
- For batch:
p-limitpackage (npm install p-limit)
Instructions
Step 1: Transcription Service Class
import { createClient, DeepgramClient } from '@deepgram/sdk';
import { readFileSync } from 'fs';
interface TranscribeOptions {
model?: 'nova-3' | 'nova-2' | 'nova-2-meeting' | 'nova-2-phonecall' | 'base';
language?: string;
diarize?: boolean;
utterances?: boolean;
paragraphs?: boolean;
smart_format?: boolean;
summarize?: boolean; // Audio intelligence
detect_topics?: boolean; // Topic detection
sentiment?: boolean; // Sentiment analysis
intents?: boolean; // Intent recognition
keywords?: string[]; // Keyword boosting: ["term:weight"]
callback?: string; // Async callback URL
}
class DeepgramTranscriber {
private client: DeepgramClient;
constructor(apiKey: string) {
this.client = createClient(apiKey);
}
async transcribeUrl(url: string, opts: TranscribeOptions = {}) {
const { result, error } = await this.client.listen.prerecorded.transcribeUrl(
{ url },
{
model: opts.model ?? 'nova-3',
language: opts.language ?? 'en',
smart_format: opts.smart_format ?? true,
diarize: opts.diarize ?? false,
utterances: opts.utterances ?? false,
paragraphs: opts.paragraphs ?? false,
summarize: opts.summarize ? 'v2' : undefined,
detect_topics: opts.detect_topics ?? false,
sentiment: opts.sentiment ?? false,
intents: opts.intents ?? false,
keywords: opts.keywords,
callback: opts.callback,
}
);
if (error) throw new Error(`Transcription failed: ${error.message}`);
return result;
}
async transcribeFile(filePath: string, opts: TranscribeOptions = {}) {
const audio = readFileSync(filePath);
const mimetype = this.detectMimetype(filePath);
const { result, error } = await this.client.listen.prerecorded.transcribeFile(
audio,
{
model: opts.model ?? 'nova-3',
smart_format: opts.smart_format ?? true,
mimetype,
diarize: opts.diarize ?? false,
utterances: opts.utterances ?? false,
summarize: opts.summarize ? 'v2' : undefined,
detect_topics: opts.detect_topics ?? false,
sentiment: opts.sentiment ?? false,
}
);
if (error) throw new Error(`File transcription failed: ${error.message}`);
return result;
}
private detectMimetype(path: string): string {
const ext = path.split('.').pop()?.toLowerCase();
const map: Record<string, string> = {
wav: 'audio/wav', mp3: 'audio/mpeg', flac: 'audio/flac',
ogg: 'audio/ogg', m4a: 'audio/mp4', webm: 'audio/webm',
};
return map[ext ?? ''] ?? 'audio/wav';
}
}
Step 2: Extract Structured Results
function formatResult(result: any) {
const channel = result.results.channels[0];
const alt = channel.alternatives[0];
return {
transcript: alt.transcript,
confidence: alt.confidence,
words: alt.words?.map((w: any) => ({
word: w.word,
start: w.start,
end: w.end,
confidence: w.confidence,
speaker: w.speaker, // Only if diarize: true
punctuated_word: w.punctuated_word,
})),
// Speaker segments (requires utterances: true + diarize: true)
utterances: result.results.utterances?.map((u: any) => ({
speaker: u.speaker,
text: u.transcript,
start: u.start,
end: u.end,
confidence: u.confidence,
})),
// Audio intelligence results
summary: result.results.summary?.short,
topics: result.results.topics?.segments,
sentiments: result.results.sentiments?.segments,
intents: result.results.intents?.segments,
metadata: {
duration: result.metadata.duration,
channels: result.metadata.channels,
model: result.metadata.model_info,
request_id: result.metadata.request_id,
},
};
}
Step 3: Batch Processing
import pLimit from 'p-limit';
async function batchTranscribe(
files: string[],
opts: TranscribeOptions = {},
concurrency = 5
) {
const transcriber = new DeepgramTranscriber(process.env.DEEPGRAM_API_KEY!);
const limit = pLimit(concurrency);
const results = await Promise.allSettled(
files.map(file =>
limit(async () => {
const result = await transcriber.transcribeFile(file, opts);
console.log(`Done: ${file} (${result.metadata.duration}s)`);
return { file, result: formatResult(result) };
})
)
);
const succeeded = results.filter(r => r.status === 'fulfilled');
const failed = results.filter(r => r.status === 'rejected');
console.log(`Batch complete: ${succeeded.length} ok, ${failed.length} failed`);
return results;
}
Step 4: Async Callback Transcription (Large Files)
// For files >2 hours or when you don't want to hold a connection open,
// use Deepgram's callback feature. Deepgram POSTs results to your URL.
async function submitAsync(audioUrl: string, callbackUrl: string) {
const transcriber = new DeepgramTranscriber(process.env.DEEPGRAM_API_KEY!);
// Deepgram returns a request_id immediately, processes in background
const result = await transcriber.transcribeUrl(audioUrl, {
model: 'nova-3',
diarize: true,
callback: callbackUrl, // Your HTTPS endpoint
});
console.log('Submitted. Request ID:', result.metadata.request_id);
// Deepgram will POST results to callbackUrl when done
// Retries up to 10 times with 30s delay on failure
}
Step 5: Keyword Boosting
// Boost domain-specific terms for higher accuracy
const result = await transcriber.transcribeUrl(audioUrl, {
model: 'nova-3',
keywords: [
'Kubernetes:1.5', // Boost weight 1.0-2.0
'PostgreSQL:1.5',
'microservices:1.3',
],
});
Output
DeepgramTranscriberclass with URL and file transcription- Structured result extraction with word-level timing, speakers, and intelligence
- Batch processing with configurable concurrency via
p-limit - Async callback pattern for large files
- Keyword boosting for domain vocabulary
Error Handling
| Error | Cause | Solution |
|---|---|---|
400 Bad Request |
Invalid audio format | Verify file header bytes (WAV: RIFF, MP3: 0xFFF3/0xFFFB) |
413 Payload Too Large |
File exceeds limit | Use callback URL for async processing |
| Empty transcript | No speech in audio | Check audio volume, try alternatives: 3 for confidence |
408 Timeout |
Long file, sync mode | Switch to callback-based async |
| Low confidence | Background noise | Preprocess: ffmpeg -i input.wav -af "highpass=f=200,lowpass=f=3000" clean.wav |
Resources
Next Steps
Proceed to deepgram-core-workflow-b for real-time streaming transcription.
04 同仓库更多 Skills
zustand-store-creator
Create zustand store creator operations. Auto-activating skill for Frontend Development. Triggers on: zustand store creator, zustand store creator Part of the Frontend Development skill category. Use when working with zustand store creator functionality. Trigger with phrases like "zustand store creator", "zustand creator", "zustand".
zero-trust-config-helper
Configure with zero trust config helper operations. Auto-activating skill for Security Advanced. Triggers on: zero trust config helper, zero trust config helper Part of the Security Advanced skill category. Use when configuring systems or services. Trigger with phrases like "zero trust config helper", "zero helper", "zero".
zapier-integration-helper
Assist with zapier integration helper operations. Auto-activating skill for Business Automation. Triggers on: zapier integration helper, zapier integration helper Part of the Business Automation skill category. Use when working with APIs or building integrations. Trigger with phrases like "zapier integration helper", "zapier helper", "zapier".
yaml-config-validator
Validate yaml config validator operations. Auto-activating skill for DevOps Basics. Triggers on: yaml config validator, yaml config validator Part of the DevOps Basics skill category. Use when configuring systems or services. Trigger with phrases like "yaml config validator", "yaml validator", "yaml".
xvary-stock-research
Thesis-driven equity analysis from public SEC EDGAR and market data; /analyze, /score, /compare workflows with bundled Python tools (Claude Code, Cursor, Codex).
xss-vulnerability-scanner
Scan xss vulnerability scanner operations. Auto-activating skill for Security Fundamentals. Triggers on: xss vulnerability scanner, xss vulnerability scanner Part of the Security Fundamentals skill category. Use when working with xss vulnerability scanner functionality. Trigger with phrases like "xss vulnerability scanner", "xss scanner", "xss".