Build a Voice‑AI Portfolio That Gets Recruiters Talking
You’ve spent months mastering speech synthesis, building a custom voice clone, and polishing your code. Now you need a portfolio that showcases those skills and actually lands interviews. A live, interactive demo site is the most convincing proof‑point: recruiters can hear the voice in real time, inspect the code, and see you’re comfortable with the whole stack. Below is a practical roadmap that covers everything from data preparation to deployment—plus a couple of tools that will make the process painless.
1. Gather the Right Data
A good voice clone starts with clean, high‑quality recordings. Aim for at least 10 minutes of speech in a quiet environment, recorded at 44.1 kHz. Use a pop‑filter and a condenser mic if possible. Here’s a quick checklist:
| Item | Why it matters |
|---|---|
| Consistent pacing | Keeps the model from learning random pauses. |
| Balanced phoneme coverage | Helps the model synthesize uncommon words. |
| Minimal background noise | Reduces artifacts in the output. |
Once you have your raw files, run a quick script to trim silences and normalize volume. The following Python snippet uses pydub:
from pydub import AudioSegment
import glob, os
def clean_audio(file_path, output_dir):
audio = AudioSegment.from_file(file_path)
trimmed = audio.strip_silence(silence_len=500, silence_thresh=-50)
trimmed.normalize().export(os.path.join(output_dir, os.path.basename(file_path)), format="wav")
for f in glob.glob("raw/*.wav"):
clean_audio(f, "clean")
Save the cleaned clips in a folder named clean. Those will be the training set you’ll upload to ElevenLabs.
2. Clone the Voice with ElevenLabs
ElevenLabs offers a straightforward API for training a custom voice and generating speech. Sign up (free tier is generous) and create a new voice using your clean folder. The API expects a ZIP of WAV files.
a. Create the Voice
curl -X POST "https://api.elevenlabs.io/v1/voices" \
-H "xi-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"name": "John Doe",
"files": ["clean/part1.wav","clean/part2.wav"]
}'
Tip – The API accepts up to 1 GB of data. If your dataset is larger, split it into multiple uploads and merge them in the dashboard.
b. Generate Speech
import requests, json
API_KEY = "YOUR_API_KEY"
VOICE_ID = "1234567890abcdef"
def synthesize(text, voice_id=VOICE_ID):
url = f"https://api.elevenlabs.io/v1/text-to-speech/{voice_id}"
headers = {
"xi-api-key": API_KEY,
"Content-Type": "application/json"
}
payload = {"text": text, "voice_settings": {"stability":0.5,"similarity_boost":0.75}}
resp = requests.post(url, headers=headers, json=payload)
resp.raise_for_status()
with open("output.mp3", "wb") as f:
f.write(resp.content)
synthesize("Hello, this is your custom voice demo!")
The response is an MP3 you can embed directly into your site.
3. Build the Frontend
A clean, minimal UI keeps the focus on the voice. Use vanilla JS or a lightweight framework like Svelte. Below is a vanilla example that pulls a short script from the backend and plays it.
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Voice AI Demo</title>
</head>
<body>
<h1>Talk to My Voice</h1>
<textarea id="script" rows="4" cols="50">Say something to hear it back.</textarea><br>
<button id="speak">Speak</button>
<audio id="player" controls></audio>
<script>
document.getElementById('speak').onclick = async () => {
const text = document.getElementById('script').value;
const res = await fetch('/api/synthesize', {method:'POST', headers:{'Content-Type':'application/json'}, body: JSON.stringify({text})});
const blob = await res.blob();
document.getElementById('player').src = URL.createObjectURL(blob);
};
</script>
</body>
</html>
Your /api/synthesize endpoint simply forwards the request to ElevenLabs and streams the MP3 back. That keeps the key hidden on the server side.
4. Set Up the Backend
A minimal Express server in Node.js is enough. Install the dependencies:
npm init -y
npm install express axios dotenv
Create .env with your ElevenLabs API key:
ELEVENLABS_KEY=YOUR_API_KEY
And index.js:
require('dotenv').config();
const express = require('express');
const axios = require('axios');
const app = express();
app.use(express.json());
app.use(express.static('public')); // your HTML + JS
app.post('/api/synthesize', async (req, res) => {
const { text } = req.body;
try {
const response = await axios.post(`https://api.elevenlabs.io/v1/text-to-speech/1234567890abcdef`, {
text,
voice_settings: { stability: 0.5, similarity_boost: 0.75 }
}, {
headers: {
'xi-api-key': process.env.ELEVENLABS_KEY,
'Content-Type': 'application/json'
},
responseType: 'arraybuffer'
});
res.set('Content-Type', 'audio/mpeg');
res.send(response.data);
} catch (err) {
console.error(err);
res.status(500).send('Error generating speech');
}
});
app.listen(3000, () => console.log('Server running on http://localhost:3000'));
Run it:
node index.js
Open http://localhost:3000 in your browser and test the demo.
5. Deploy with Bluehost
You can host the entire stack on Bluehost, which offers a free domain for the first year and affordable shared hosting. The platform supports Node.js and has a simple one‑click deployment wizard.
- Sign up – Use the affiliate link: https://bluehost.sjv.io/5k0d52
- Create a new Node app – In your cPanel, go to Node.js > Setup Application. Point the root to your project folder.
- Upload files – Use the File Manager or an FTP client.
-
Install dependencies – In the Node app console, run
npm install. -
Start the app – Set the startup file to
index.js.
Bluehost’s support team is very responsive, and the uptime is reliable for a portfolio site. The cost is a fraction of what you’d pay for a dedicated server, and you get a free SSL certificate, which is a must for any demo that involves audio streaming.
6. Polish and Showcase
- Add a README – Explain the project, the tech stack, and the results.
- Include a GitHub repo – Recruiters love to see your commit history.
- Embed the demo on your LinkedIn or personal blog.
- Add a “Download” button – Let visitors download the MP3s you generate.
- Measure engagement – Add simple analytics (e.g., Google Analytics) to see how many people are interacting with the demo.
Final Thoughts
You’ve built a voice‑AI demo that:
- Uses clean data and a robust TTS engine (ElevenLabs).
- Demonstrates end‑to‑end flow: data → model → API → UI.
- Is hosted on an affordable, easy‑to‑use platform (Bluehost).
All that remains is to share your project on dev.to, GitHub, and LinkedIn. Recruiters will notice that you can take a voice‑AI idea from concept to a live, interactive demo—exactly the skill set they’re looking for.
Ready to Show Off Your Voice AI?
Try ElevenLabs today and let your voice speak for you: https://try.elevenlabs.io/kr07zfuqn1bp
And deploy your portfolio effortlessly with Bluehost: https://bluehost.sjv.io/5k0d52
Good luck, and happy coding!
Top comments (0)