Tapestry is using GenAI on AWS to centralize knowledge across business units, making company information more accessible and actionable.
Tapestry implemented AWS’ GenAI knowledge management system to overcome data silos across IT, HR, and legal teams. The GenAI solution is built on Amazon Bedrock, utilizing Claude 3 Haiku for natural language processing and Amazon Titan foundation models for text embeddings. Aurora PostgreSQL serves as the vector database, indexing millions of document chunks to allow fast and accurate retrieval.
Employees can query the system through a chatbot interface hosted on Amazon CloudFront, which connects them to relevant knowledge stored in Amazon S3. AWS Lambda functions orchestrate search queries, applying security protocols via AWS Identity and Access Management to ensure appropriate access control. By centralizing corporate knowledge from IT, HR, and legal teams, the AI system eliminates the delays caused by fragmented information storage.
Tapestry has also implemented automated processes to keep knowledge bases updated. AI-powered pipelines scan and process new or modified documents. Generating fresh embeddings and ensuring information remains current. The scalable, serverless architecture allows employees worldwide to access real-time knowledge. Reducing dependency on subject matter experts and improving response times to critical inquiries. By minimizing time spent searching for answers, the AI system enables employees to focus on strategic initiatives. Rather than administrative tasks.
Future AI-driven enhancements include voice search, image-based queries, and structured data analytics for real-time insights on sales, inventory, and store traffic. By embedding GenAI into enterprise workflows, Tapestry is not only streamlining knowledge management but also fostering a culture of innovation and agility. The company sees AI as a transformative force in reshaping corporate decision-making. Ensuring that employees can access accurate, relevant information faster than ever before.