Inspiration
Amazon sellers often struggle with poor product visibility and low conversions due to unoptimized listings and weak visuals. We wanted to create an autonomous AI agent that optimizes both SEO content and product media, helping sellers compete effectively without hiring expensive marketing agencies.
What it does
Amazon Product Optimizer is a cloud-native SaaS platform that empowers Amazon sellers to: Log in securely and analyze product listings Compare SEO performance with competitors Generate optimized titles, bullet points, and descriptions using AI Enhance product images Benchmark SEO performance All of this is orchestrated through AWS Bedrock AgentCore, which acts as the brain of the system — executing SEO, media, and analysis tasks autonomously.
How we built it
Frontend: ReactJS hosted on AWS Amplify, with a sleek UI for product input and result visualization. Auth Layer: AWS Cognito handles secure seller logins and token-based access. Backend: EC2 With Agents: Product scraping and metadata extraction Competitor keyword analysis AI-powered SEO generation via Bedrock (Titan & Claude models) Image enhancement using Bedrock Image APIs Core AI Logic: AgentCore coordinates all workflows — triggering the right AI model for text, image, or video enhancement.
Challenges we ran into
Building a scalable multi-agent pipeline that manages text, image, and video generation workflows efficiently. Maintaining low latency for SEO content generation while using large foundation models. Designing a universal scraper for various Amazon product formats and regional sites. Managing secure access tokens and rate limits for multiple seller accounts.
Accomplishments that we're proud of
Fully functional AI SEO pipeline that can optimize any Amazon listing in under a minute. End-to-end AWS-native integration — from login to AI generation — built entirely on serverless tech. Seamless orchestration through AgentCore, showcasing how autonomous AI agents can perform structured e-commerce automation. Clean, intuitive frontend that even non-technical sellers can use effortlessly.
What we learned
How to integrate multiple AI modalities (text, image, video) under one orchestration agent. The importance of prompt design and grounding AI output for SEO accuracy and product authenticity. Best practices for serverless scalability and data privacy when handling user-generated media.
What's next for Product Optimizer – AI-Powered SEO & Media Agent
Integrating directly with Amazon Seller Central APIs to auto-update listings. Expanding the AI suite with a PPC Optimization Agent for ad performance. Introducing multi-marketplace support (eBay, Shopify). Deploying a localized version for offline sellers or regional e-commerce platforms.
Built With
- amazon-ec2
- amazon-web-services
- bedrock
- github
- llm
- python
- react
- restapi
- s3
Log in or sign up for Devpost to join the conversation.