Cyberspace has become an essential place to communicate and network with people worldwide. However, the perceived anonymity and accessibility of online platforms have created an environment that facilitates the spread of hateful content, boosting online violence, and connecting individuals who share similar hateful beliefs. We are designing and developing the foundation of transformative technology for countering hate speech in online social spaces to ultimately mitigate violence-inducing behavior. This paper introduces a conceptual framework for countering hate speech, describes the victim and stakeholder perspectives that shaped it, and outlines a number of use-case scenarios for the tool in development. For this research we focus on the context of English and Norwegian languages data. Our system is designed to handle multi-modal data, including not only textual data but also visual and auditory information. By involving the widest number of stakeholders at local, regional, national level, including potential users of the system, we conducted a comprehensive analysis of stakeholder needs. The innovative multi-lingual, multi-modal analytical framework presents directions for leveraging machine intelligence, inferred from data via state-of-the-art Artificial Intelligence techniques, while also incorporating ethical guidelines for responsible technological development.