AutoScout24, a major European automotive marketplace, has launched Bot Factory to standardize the creation and deployment of AI agents across its organization. Developed in partnership with AWS, this centralized framework moves the company beyond isolated experiments into a secure, enterprise-grade development environment. The system utilizes Amazon Bedrock to allow various teams to build specialized agents capable of reasoning, interacting with company data, and performing complex tasks using a unified set of tools.

The first production application is an internal support bot designed to assist platform engineers who previously spent 30% of their time on repetitive administrative tasks. By integrating Retrieval Augmented Generation (RAG), the bot searches internal documentation to provide accurate answers and uses secure API connections to perform actions like granting software licenses. This automation significantly reduces manual workloads while ensuring high accuracy through verified internal data sources and secure tool execution.

Technically, the Bot Factory relies on an "agents-as-tools" architecture where a central orchestrator identifies user intent and delegates workflows to specialized sub-agents. Built on a serverless infrastructure, the platform ensures cost-efficiency and session isolation to prevent data leaks between different user interactions. Utilizing the Strands Agents SDK, the framework provides a scalable blueprint for future innovations, such as automated code reviews and enhanced customer service tools, while maintaining rigorous enterprise security standards.

AutoScout24, a major European automotive marketplace, has launched Bot Factory to standardize the creation and deployment of AI agents across its organization. Developed in partnership with AWS, this centralized framework moves the company beyond isolated experiments into a secure, enterprise-grade development environment. The system utilizes Amazon Bedrock to allow various teams to build specialized agents capable of reasoning, interacting with company data, and performing complex tasks using a unified set of tools.

The first production application is an internal support bot designed to assist platform engineers who previously spent 30% of their time on repetitive administrative tasks. By integrating Retrieval Augmented Generation (RAG), the bot searches internal documentation to provide accurate answers and uses secure API connections to perform actions like granting software licenses. This automation significantly reduces manual workloads while ensuring high accuracy through verified internal data sources and secure tool execution.

Technically, the Bot Factory relies on an "agents-as-tools" architecture where a central orchestrator identifies user intent and delegates workflows to specialized sub-agents. Built on a serverless infrastructure, the platform ensures cost-efficiency and session isolation to prevent data leaks between different user interactions. Utilizing the Strands Agents SDK, the framework provides a scalable blueprint for future innovations, such as automated code reviews and enhanced customer service tools, while maintaining rigorous enterprise security standards.