Chatbots are agents that enable the interaction of users and software by means of written or spoken natural language conversation. Their use is growing, and many companies are starting to offer their services via chatbots, e.g., for booking, shopping or customer support. For this reason, many chatbot development tools have emerged, which makes choosing the most appropriate tool difficult. Moreover, there is hardly any support for migrating chatbots between tools.

To alleviate these issues, we propose a model-driven engineering solution that includes: (i) a domain-specific language to model chatbots independently of the development tool; (ii) a recommender that suggests the most suitable development tool for the given chatbot requirements and model; (iii) code generators that synthesize the chatbot code for the selected tool; and (iv) parsers to extract chatbot models out of existing chatbot implementations. Our solution is supported by a web IDE called Conga that can be used for both chatbot creation and migration. A demo video is available at https://youtu.be/3sw1FDdZ7XY.