This paper presents findings from the co-design of Svea, a Generative AI (GAI) chatbot integrated into the mobile application KomIn Sverige, designed to support foreign-born job seekers in Sweden. Addressing key challenges such as limited digital literacy, language barriers, and professional networking gaps, the purpose of Svea is to offer personalized guidance and interactive support. Through our co-design approach, involving job seekers and job coaches, we have designed, integrated and refined Svea’s functionalities to balance automation with human support, enhancing usability through simplified language and FAQ-driven prompts. Initial testing highlights the importance of iterative development in ensuring accessibility and adaptability and reveals how carefully designed affordances can support action and understanding among users operating on varying levels of digital literacy. These findings contribute to the HCI field by demonstrating how GAI chatbots can empower marginalized communities and improve employment outcomes. Future work will explore scalability, voice-enabled chat, and ethical considerations in deploying AI-driven employment support tools.