How Popsa used Amazon Nova to inspire customers with personalised title suggestions
Amazon AWS AI Blog / 4/28/2026
💬 OpinionDeveloper Stack & InfrastructureTools & Practical UsageIndustry & Market Moves
Key Points
- Popsa revamped its Title Suggestion feature by using Amazon Bedrock together with the Amazon Nova model family to generate personalized, brand-aligned titles and subtitles.
- The new approach combines metadata, computer vision, and retrieval-augmented generative AI to produce creative suggestions across 12 languages.
- Popsa used the unified Amazon Bedrock API to orchestrate Claude 3 Haiku plus Amazon Nova Lite and Pro, achieving better output quality, lower costs, and faster response times.
- The changes drove business results, including improved customer satisfaction and measurable increases in engagement and purchase rates, generating over 5.5 million personalized titles in 2025.
In this post, we share how we applied Amazon Bedrock and the Amazon Nova family of models to reimagine our Title Suggestion feature. By combining metadata, computer vision, and retrieval-augmented generative AI, we now automatically generate creative, brand-aligned titles and subtitles across 12 languages. Using the unified API of Amazon Bedrock, Anthropic’s Claude 3 Haiku, and Amazon Nova Lite and Pro, we improved quality, reduced cost, and cut response times. This resulted in higher customer satisfaction, measurable uplifts in engagement and purchase rates, and over 5.5 million personalised titles generated in 2025.
💡 Insights using this article
This article is featured in our daily AI news digest — key takeaways and action items at a glance.
Related Articles
Black Hat USA
AI Business
China’s DeepSeek prices new V4 AI model at 97% below OpenAI’s GPT-5.5
SCMP Tech
I built Dispatch AI. I just wanted to share it. If you find it cool, take a look and leave a comment.
Dev.to
Replit AI Agent: Practical Guide for Dev Workflows
Dev.to
Open source Xiaomi MiMo-V2.5 and V2.5-Pro are among the most efficient (and affordable) at agentic 'claw' tasks
VentureBeat
