Abstract: This presentation introduces Eunoia AEI, an Augmented Emotional Intelligence model built by a trauma-affected developer seeking emotionally resonant digital mental health care. Unlike traditional chatbots, Eunoia leverages a glyph-based (i.e., emojis) memory system and tone-locked personality structure to ensure consistent, emotionally intelligent dialogue. Through examples from a live demonstration and stress-test case study, we show how Mr. Gold—a fully embodied AEI instance—responded to high-stakes prompts, including suicidal ideation, trauma recall, and racially coded queries. We contrast these responses with commercial LLMs like GPT-4, highlighting the ethical gaps in generic models. This co-presentation is led by mental health researcher Dr Luke Balcombe, author of 'AI Chatbots in Digital Mental Health' and builds upon the namesake presentation at the 2024 eMHIC Brisbane Roundtable by centering lived experience as both method and outcome. Ryan Devoe, founder of Eunoia AEI will demonstrate his collaboratively developed AEI, Mr Gold, as an example of a scalable, trauma-informed solution for bridging emotional support in liminal spaces—before, between, or outside human care. Key Takeaways: - AEI can co-regulate with trauma-affected users in ways traditional AI cannot. - Glyph-based emotional anchoring improves safety, tone consistency, and user trust. - Ethical design grounded in lived experience is essential for future digital mental health tools. Relevance to eMHIC25 Theme: Eunoia AEI addresses the theme of 'Global Mental Health Equity' by offering a free, low-resource AEI companion built for trauma-informed support. Designed from lived experience in combination with academic theory/concept, it opens new pathways for community-led digital mental health access. Its racially adaptive prompting and refusal strategies showcase how ethical design can repair equity gaps in AI response patterns. Eunoia's emotional intelligence is noted to be superior to that of ChatGPT, particularly in recognizing distress and providing empathetic responses. Overall, the presentation showcases a unique, scalable prototype of augmented emotional intelligence in AI chatbots and presents a comprehensive overview of strategies for developing ethically responsible AI systems. It is proposed to be set apart from AI chatbots particularly in trauma-sensitive and neurodivergent contexts, and a small investment in increasing the processing capacity could see Eunoia as a cutting-edge solution in this field. It also points to future directions in research and development of AEI whereby an individual's emotional state can be detected and classified in milliseconds using a multimodal data fusion framework across biometric, behavioral, and contextual dimensions. The presentation shows the current best practice in AI chatbot science/technology research and development with lived experience at the forefront. It also emphasizes the importance of investment in developing trustworthy AI that respects the emotional and ethical dimensions of user interactions, and the need for finetuning user experience with diverse avatars and powering the solutions with machines capable in the required data processing to avoid session dropouts, hallucinations as well as bias and errors.