In this paper, we investigate how to design a security architecture of a Platform-as-a-Service (PaaS) solution, denoted as Secure Virtual Premise (SVP), for collaborative and distributed AI engineering using AI artifacts and Machine Learning (ML) pipelines. Artifacts are re-usable software objects which are a) tradeable in marketplaces, b) implemented by containers, c) offer AI functions as microservices, and, d) can form service chains, denoted as AI pipelines. Collaborative engineering is facilitated by the trading and (re-)using artifacts and, thus, accelerating the AI application design.

The security architecture of the SVP is built around the security needs of collaborative AI engineering and uses a proxy concept for microservices. The proxy shields the AI artifact and pipelines from outside adversaries as well as from misbehaving users, thus building trust among the collaborating parties. We identify the security needs of collaborative AI engineering, derive the security challenges, outline the SVP’s architecture, and describe its security capabilities and its implementation, which is currently in use with several AI developer communities. Furthermore, we evaluate the SVP’s Technology Readiness Level (TRL) with regard to collaborative AI engineering and data security.

When exposed to the network, applications and devices are exposed to constant security risks. This puts pressure on hardware and software vendors to test even more than before how secure applications and devices are before being released to customers.

We have worked towards defining and developing a frame- work for automated security testbeds. Testbeds comprise both the ability to build on-demand virtual isolated networks that emulate corporate networks, as well as the ability to automate security breach scenarios, which accelerates the testing process. In order to accomplish both features of the testbed, we have based the framework on well-established cloud and orchestration technologies e. g. , OpenStack and Ansible. Although many of these technologies are powerful, they are also complex, leading to a steep learning curve for new users. Thus, one of the main goals of the developed framework is to hide the underlying complexities through a template approach and a simplified user interface that shortens the initial training time.

In this paper, we present the full stack of technologies that were used for constructing the testbed framework. The framework allows us to create entire virtual networks and to manipulate network devices started in it, via comprehensive yet simple interfaces. Also, we describe a specific testbed solution, developed as a part of the Test Arena Blekinge project.

Service Chains have developed into an important concept in service provisioning in today’s and future Clouds. Cloud systems, e.g., Amazon Web Services (AWS), permit the implementation and deployment of new applications, services and service chains rapidly and flexibly. They employ the idea of Infrastructure as Code (IaC), which is the process of managing and provisioning computing infrastructure and its configuration through machine-processable definition files.

In this paper, we first detail future service chains with particular focus on Network Function Virtualization (NFV) and machine learning in AI. Afterwards, we analyze and summarize the capabilities of today’s IaC tools for orchestrating Cloud infrastructures and service chains. We compare the functionality of the major five IaC tools: Puppet, Chef, SaltStack, Ansible, and Terraform. In addition, we demonstrate how to analyze the functional capabilities of one of the tools. Finally, we give an outlook on future research issues on using IaC tools across multiple operators, data center domains, and different stockholders that collaborate on service chains.