Walkthrough Complete - Text-to-Speech
You’ve completed the Text-to-Speech walkthrough
Great! You've deployed the demo
Now let's walk through what you just deployed and see it in action.
Start WalkthroughChoose your next step
Generate Evidence Pack
Create your business case documentation with what you've learned.
Generate Evidence PackWalkthrough complete
Congratulations! In just 8 minutes, you've seen how Amazon Polly generates broadcast-quality audio from text at 99.99% lower cost than studio recording while ensuring full accessibility compliance. What would have taken weeks to pilot with voice talent, you've experienced hands-on right now.
What you've learned
Natural-sounding neural voices
You heard Amy's conversational warmth, Brian's authoritative clarity, and Emma's formal professionalism. These aren't robotic "computer voices" - Polly's neural engine generates speech with natural inflection, breathing patterns, and emotional tone appropriate for each announcement type.
Value: Residents won't know it's AI-generated - quality rivals professional studios
Instant generation vs weeks of delay
Text-to-speech converts announcements to audio in 30 seconds. No booking voice talent, no studio scheduling, no editing rounds. Paste text, select voice, click generate. If the announcement text changes, regenerate audio instantly for £0.008 instead of £100 re-recording fee.
Value: 99.98% faster (1-2 weeks to 30 seconds) with zero coordination overhead
99.99% cost reduction
Studio recording costs £150-£300 per announcement (talent, studio time, editing). Polly costs £0.008 per typical announcement. For 50 announcements annually, savings of £10,000 - £15,000 can be reallocated to frontline services.
Value: Pay-per-use pricing with zero upfront investment - low risk, high return
Full accessibility compliance
Text-to-speech ensures every announcement has an audio version, meeting WCAG 2.1 AA requirements under PSBAR 2018. Serves 2% of residents with visual impairments, 16% with literacy difficulties, and 19% aged 65+. Avoids penalties up to £20,000 per accessibility breach.
Value: Legal compliance + inclusive service delivery + reputational protection
Committee-ready talking points
Use these points when presenting to decision-makers:
The problem
"Currently, council announcements are published as text only on our website and printed notices. This excludes 2% of residents with visual impairments (1.3M UK residents), 16% with literacy below Level 1, and 19% aged 65+ who may have age-related vision decline. PSBAR 2018 requires accessible alternatives, but studio recording costs £150-£300 per announcement with 1-2 week turnaround. Due to cost and time constraints, we rarely produce audio versions, creating compliance risk under Equality Act 2010 with potential penalties up to £20,000 per breach."
The solution
"Amazon Polly text-to-speech generates broadcast-quality audio from announcement text in 30 seconds. UK English neural voices (Amy, Brian, Emma) sound natural and professional, not robotic. Cost is £0.016 per 1000 characters = £0.008 per typical announcement. Communications officers paste text, select voice, generate audio - 1 minute total. Audio file is saved to website and can be integrated with phone systems for automated announcements. No infrastructure required - pay only for audio generated."
The value
"For 50 announcements annually: Annual cost savings of £9,999.60 (from £10,000 studio recording to £0.40 Polly usage). Time savings of 99 hours per year (from 120 min coordination per announcement to 1 min). Compliance benefit: 100% WCAG 2.1 AA compliance avoiding up to £20K penalties per breach. Accessibility benefit: Multi-channel delivery (web, phone, social media) ensures information reaches all residents regardless of ability. Reputational benefit: Demonstrates council commitment to digital inclusion and modern service delivery."
The risk
"Very low. Pay-per-use pricing means no upfront investment. £0.40 annual cost for 50 announcements is negligible (less than one coffee). Trial with 10 announcements costs £0.08 to validate voice quality and resident acceptance. Polly is a managed AWS service requiring zero infrastructure or maintenance. Integrates with existing CMS (WordPress, Drupal, Joomla) via simple API. Typical pilot: 1-2 months with 10-20 announcements to gather resident feedback and refine voice selection."
Clean up your resources
Good news: Your resources will automatically delete after 2 hours from deployment. However, you can delete them now to stop any further charges immediately.
- Database content (RDS Aurora)
- Uploaded files (EFS storage)
- Any configuration changes you made
Step-by-step deletion
- Open the CloudFormation console
-
Find your stack
Look for a stack named:
ndx-try-text-to-speech-[timestamp]The timestamp is when you deployed. You can sort by "Created time" to find recent stacks.
-
Select and delete
Select the checkbox next to your stack, then click the Delete button.
-
Confirm deletion
Click Delete in the confirmation dialog. The stack status will change to
DELETE_IN_PROGRESS. -
Wait for completion
Deletion typically takes 5 to 10 minutes. The stack will disappear from the list when complete.
Costs stop after deletion
Estimated evaluation cost: Less than $0.50 for a 15-minute trial
Troubleshooting
Stack shows DELETE_FAILED status
This usually happens when resources can't be automatically cleaned up. Common causes:
- S3 bucket not empty: The bucket may contain files. Go to S3, empty the bucket manually, then retry deletion.
- Lambda functions in use: Wait a few minutes and retry. Sometimes functions take time to fully stop.
- Network interfaces still attached: These usually clear within 5 to 10 minutes. Retry the deletion.
To retry deletion: Select the failed stack and click Delete again.
I can't find my stack in the list
If your stack isn't visible:
- Check the region: Make sure you're viewing US East (N. Virginia) in the console header.
- Stack already deleted: It may have auto-deleted after 2 hours. No action needed!
- View deleted stacks: Click "View nested" dropdown and select "Deleted" to see recently deleted stacks.
Extending your evaluation time
If you want to continue testing beyond the 2-hour limit:
- Resources will auto-delete after 2 hours total (from deployment)
- Cost: approximately $1-2 per hour of active testing
- Maximum cost is capped by template configuration
You can redeploy the scenario anytime to start fresh with a new 2-hour window.
Still having trouble? Contact the NDX:Try team or report an issue on GitHub (opens in new tab).
Production considerations
If you're considering deploying text-to-speech for real council announcements:
Integration with council CMS (WordPress, Drupal, etc.)
Production deployment integrates with your content management system:
- CMS plugin auto-generates audio when announcement is published
- Embed HTML5 audio player automatically on announcement page
- Store audio files in S3 with CloudFront CDN for fast global delivery
- Optional: API endpoint allows manual audio generation for non-CMS content
- Webhook triggers can regenerate audio if announcement text is updated
Phone system (IVR) integration for automated announcements
Integrate Polly audio with council phone systems:
- Amazon Connect (cloud contact center) has native Polly integration
- Legacy IVR systems: Upload MP3 files to phone system via FTP/API
- Dynamic menus: Generate audio for phone prompts on demand
- Emergency hotlines: Update recorded messages instantly during crises
- Multilingual support: Generate announcements in 29 languages for diverse communities
Handling pronunciation of council-specific terms and place names
For specialized terminology:
- Create pronunciation lexicon with custom pronunciation rules for your council
- Upload lexicon to Polly - applies to all future audio generation automatically
- Use SSML (Speech Synthesis Markup Language) for inline pronunciation overrides
- Example: <phoneme alphabet='ipa' ph='rɛdɪŋ'>Reading</phoneme> for town name
- Most UK place names, postcodes, street names pronounced correctly by default
- Test and refine pronunciation during pilot phase
Cost at scale and budget planning
Real-world cost examples (neural voices, 500-character announcements):
- 50 announcements/year: ~£0.40/year (£0.008 per announcement)
- 100 announcements/year: ~£0.80/year
- 500 announcements/year: ~£4.00/year (includes service updates, newsletters, meeting summaries)
Budget safety: Set CloudWatch billing alarm at £50/month as absolute maximum. Actual usage unlikely to exceed £1/month even with high announcement volume.
ROI example: For 100 announcements/year, £0.80 cost vs £20,000 studio recording = £19,999.20 savings + £20K compliance protection = £40K total annual value.
Data protection and GDPR compliance
Security and privacy considerations:
- Polly processes text transiently - does not store announcement content
- Audio files stored in your S3 bucket with encryption at rest (AES-256)
- Access control via IAM roles - only authorized staff can generate audio
- CloudTrail logging tracks all Polly API calls for audit compliance
- UK data residency - process all announcements in us-east-1 (N. Virginia region)
- GDPR-compliant data processing agreement with AWS
- No personally identifiable information sent to Polly (announcements are public)
Pilot phase and rollout strategy
Recommended phased approach:
- Pilot (Month 1-2): Generate audio for 10-20 announcements, embed on website, survey residents for feedback on voice quality and usefulness. Budget: £0.20
- Integration (Month 3): Integrate Polly with council CMS, train communications team, create pronunciation lexicon. Apply to all new announcements going forward. Budget: £0.50
- Backfill (Month 4): Generate audio for top 50 most-viewed historical announcements. Budget: £0.40
- Expand (Month 5-6): Extend to committee minutes, consultations, newsletters. Integrate with phone system IVR. Budget: £1.00
Total pilot cost: £2.10 over 6 months. Expected annual savings once fully deployed: £10K-£30K depending on announcement volume.
Questions or feedback?
We'd love to hear about your experience with this walkthrough: