BACK
Sellergeni – Migrating Amazon Product Listing Creation to Generative AI on AWS
Manual inspection causing costly production delays
Sellergeni partnered with Aivar to migrate their Amazon product listing platform to AWS, leveraging Bedrock Agents, Lambda, API Gateway, S3, and CloudWatch. The GenAI-powered serverless architecture automates creation and validation of product titles, descriptions, bullet points, and keywords—delivering 80% faster generation, 70% higher accuracy, full scalability.
CUSTOMER CHALLENGE
Sellergeni, serving a broad e-commerce customer base, previously relied on an on-premises generative AI model for Amazon product listing creation. To keep pace with business growth and leverage advanced generative AI capabilities, Sellergeni needed to migrate their existing LLM model to a robust, cloud-native AWS GenAI platform—enabling seamless, high-quality, and scalable product listing generation for all its sellers.
SOLUTION
Aivar partnered with Sellergeni to help migrate their automated Amazon Product Listing Builder to a cloud based GenAI-powered platform that streamlines content creation for e-commerce sellers. The solution leverages Amazon Bedrock for generative content, with Anthropic's Claude Sonnet powering creation of product descriptions and validation agents which provide the necessary guardrails. By using API Gateway for secure user integration, Amazon S3 for storage, and CloudWatch for monitoring, it fully automates the generation of product titles, descriptions, bullet points, and keywords from raw catalog data. By integrating advanced AI workflows, Aivar delivered a listing builder that not only accelerates time-to-market but also enforces Amazon’s listing best practices and consistency at scale.
Architecture
- Validated Scalability: Successfully validated the application’s ability to handle 10,000 TPS on AWS—a 3.3x increase over legacy performance.
- Operational Excellence: Achieved fully automated, drift-free deployments, reducing the Lead Time for Changes from 5 days to under 4 hours.
- Guaranteed Availability: Established a resilient architecture meeting the critical 99.9% AWS Infrastructure SLA, ensuring less than 9 hours of potential annual downtime compared to previous multi-day outages.
- Cost Efficiency: Optimized cloud spend by 25% through the use of high-performance AMD-based Amazon EC2 instances, graviton and rightsizing EKS clusters.
- Improved Deployment Velocity: Increased deployment frequency from bi-weekly releases to on-demand daily deployments using GitOps and automated Helm charts.
By implementing these AWS DevOps best practices, the partner transformed legacy manual processes into automated, high-scale cloud-native environments. These initiatives successfully reduced operational costs by up to 60% while simultaneously increasing release velocity and system reliability.
KEY OUTCOMES
Sellergeni’s GenAI-powered Amazon Product Listing Builder achieved measurable gains for e-commerce operations:
- 80% faster processing: Average listing generation time dropped from several minutes to under 15 seconds, enabling near real-time product launches.
- 70% higher accuracy: Automated generative and validation workflows improved metadata quality and compliance.
- Scalability: The architecture supports high-volume catalog operations, accommodating surges in new product uploads and onboarding more sellers efficiently.
- Seamless Migration: Successfully transitioned from the on-premise LLM model to a fully managed, cloud-native AWS generative AI platform, minimizing disruption for Sellergeni’s e-commerce customers.
.svg)
.svg)
.svg)
.svg)
.webp)
