AI processing has been a big area of focus for the research community for quite some years. The growing computation capability of edge devices has allowed Computer Vision to make use of AI techniques with greater efficiency and throughput. This approach has led to the concept of distributed computing to solve AI problems at edge. Distributed computing brings many challenges along with it like complexity, high deployment cost and security. Similar problems have been resolved in the context of Cloud Computing through containerization of the applications. This thesis makes use of the same containerization techniques to provide a simple and effective method for AI developers to create distributed systems with ease. The software framework proposed in this work makes use of Kubernetes for container orchestration. Containerization allows breaking down the entire application into smaller chunks running in isolated environments which are called microservices. Usage of microservices for distributing tasks across edge devices is a contribution of this thesis. This enables pipelining the application into microservice-level stages that can run on separate edge devices. This thesis explores the configuration of heterogeneous distributed network which incorporate Kubernetes and containerization to explore the benefits of pipelining microservices of previously monolithic applications. By utilizing Kubernetes ability to automatically orchestrate a defined network this thesis also seeks to also explore the networks sustainability and scalability based around the configuration of pipelined microservices, resource availability, and the distribution of these services within the network.