On Day 3, we’ll convert recorded voice memos into text using AssemblyAI. This transcription will later feed GPT, enabling it to generate CRM-ready sales notes.
1. Why Use Transcription?
✅ Automatic Note-Taking – Salespeople speak freely, AI turns it into text.
✅ Foundation for GPT Processing – GPT can summarize and format transcriptions into CRM entries.
✅ Better Record-Keeping – Stored transcriptions become searchable meeting logs.
2. Why AssemblyAI?
- ✅ Simple REST API.
- ✅ High Accuracy for Sales Calls.
- ✅ Free Trial + Pay-as-You-Go.
Sign up at AssemblyAI and get your API Key.
3. Setting Up AssemblyAI API
✅ Step 1: Create transcriptionService.js
Create src/api/transcriptionService.js:
import axios from 'axios';
import { ASSEMBLYAI_API_KEY } from '@env';
const AssemblyAI_BASE_URL = 'https://api.assemblyai.com/v2';
export const uploadAudio = async (fileUri) => {
const response = await fetch(fileUri);
const blob = await response.blob();
const formData = new FormData();
formData.append('file', blob, 'audio.m4a');
const uploadResponse = await axios.post(`${AssemblyAI_BASE_URL}/upload`, formData, {
headers: {
'Content-Type': 'multipart/form-data',
'Authorization': ASSEMBLYAI_API_KEY,
},
});
return uploadResponse.data.upload_url;
};
export const transcribeAudio = async (audioUrl) => {
const response = await axios.post(
`${AssemblyAI_BASE_URL}/transcript`,
{ audio_url: audioUrl },
{ headers: { Authorization: ASSEMBLYAI_API_KEY } }
);
return response.data.id;
};
export const getTranscriptionResult = async (transcriptId) => {
const response = await axios.get(`${AssemblyAI_BASE_URL}/transcript/${transcriptId}`, {
headers: { Authorization: ASSEMBLYAI_API_KEY },
});
return response.data;
};
4. Updating Home Screen to Trigger Transcription
✅ Step 1: Modify HomeScreen.js
import React, { useState } from 'react';
import { View, Text, Button, StyleSheet, ActivityIndicator, Alert } from 'react-native';
import VoiceRecorder from '../components/VoiceRecorder';
import { uploadAudio, transcribeAudio, getTranscriptionResult } from '../api/transcriptionService';
export default function HomeScreen() {
const [recordingUri, setRecordingUri] = useState(null);
const [transcription, setTranscription] = useState('');
const [isLoading, setIsLoading] = useState(false);
const handleRecordingComplete = (uri) => {
setRecordingUri(uri);
Alert.alert('Recording Saved', `Saved to: ${uri}`);
};
const handleTranscribe = async () => {
try {
setIsLoading(true);
// 1. Upload audio
const audioUrl = await uploadAudio(recordingUri);
// 2. Transcribe audio
const transcriptId = await transcribeAudio(audioUrl);
// 3. Poll for results
let status = 'processing';
let transcriptData = null;
while (status === 'processing' || status === 'queued') {
transcriptData = await getTranscriptionResult(transcriptId);
status = transcriptData.status;
if (status === 'completed') {
setTranscription(transcriptData.text);
break;
} else if (status === 'failed') {
Alert.alert('Transcription Failed');
break;
}
await new Promise((resolve) => setTimeout(resolve, 3000)); // Poll every 3s
}
} catch (error) {
console.error('Transcription Error:', error);
Alert.alert('Error', 'Failed to transcribe audio');
} finally {
setIsLoading(false);
}
};
return (
<View style={styles.container}>
<Text style={styles.title}>AI Sales Assistant</Text>
<VoiceRecorder onRecordingComplete={handleRecordingComplete} />
{recordingUri && <Button title="Transcribe Memo" onPress={handleTranscribe} />}
{isLoading && <ActivityIndicator size="large" color="#0000ff" />}
{transcription ? <Text>Transcription: {transcription}</Text> : null}
</View>
);
}
const styles = StyleSheet.create({
container: { flex: 1, justifyContent: 'center', alignItems: 'center', padding: 10 },
title: { fontSize: 24, fontWeight: 'bold', marginBottom: 20 },
});
5. Running the App
✅ Step 1: Start the App
npx expo start
✅ Step 2: Test Transcription Flow
| Action | Expected Result |
|---|---|
| Record Voice Memo | Memo recorded and URI displayed. |
| Tap Transcribe Memo | Upload → Process → Transcribed text displayed. |
| Invalid File/Slow Upload | Displays error or “Transcription Failed.” |
6. Optional: Show Status Updates
<Text>Status: {status}</Text>
7. Preparing for Tomorrow: GPT Integration
Tomorrow, we’ll:
- Send the transcription text to GPT.
- Generate summarized meeting notes and CRM action items.
8. Key Concepts Covered
✅ Integrated AssemblyAI transcription API.
✅ Uploaded voice memo files for processing.
✅ Displayed transcription result in-app.
9. Next Steps: Generating CRM Notes with GPT
Tomorrow, we’ll:
- Pass transcription text to GPT-4.
- Generate meeting summaries & CRM entries.
10. References & Learning Resources
11. SEO Keywords:
React Native voice transcription, AssemblyAI audio transcription API, React Native GPT sales assistant, voice memo to text app, AI-powered sales note automation.
