IskayPet Implements Amazon Bedrock Managed Knowledge Bases for Corporate Knowledge Unification

The Challenge of Dispersed Knowledge

IskayPet, a pet care company operating across Spain, Portugal, and France, faced a significant challenge with dispersed corporate knowledge. Operational information, such as return policies, veterinary protocols, and onboarding steps, was stored in hundreds of Microsoft SharePoint documents across four languages. Employees spent considerable time manually searching these documents to find answers to routine questions, often encountering outdated or incorrect information.

Solution with Amazon Bedrock Managed Knowledge Bases

To address this, IskayPet implemented a corporate knowledge assistant built on Amazon Bedrock Managed Knowledge Bases. This solution utilizes Retrieval Augmented Generation (RAG) without requiring IskayPet to manage vector databases, embedding pipelines, or ingestion jobs directly. The assistant was developed using Kiro with Claude Haiku 4.5.

Scope and Functionality

The new assistant provides answers to employee questions across various business areas, including retail operations, veterinary care, human resources, and procurement. It functions across four languages (Spanish, Portuguese, French, and English) and three countries, centralizing information that was previously fragmented. The system routes questions to the appropriate knowledge base, ensuring relevant and accurate responses.

Impact on Operations

By unifying its corporate knowledge, IskayPet aims to improve operational efficiency and accuracy for its 4,500 employees. The solution eliminates the time-consuming manual search process and provides a reliable way to query corporate knowledge, which is crucial as the company continues to expand its operations and add new procedures or locations.