Belong.Life developed Dave, its AI cancer mentor, with direct input from real patients, caregivers, and oncologists, using millions of anonymized data points from real-world interactions.
Behind-the-scenes look with CEO & Co Founder Eliran Malki:
1. How did you ensure Dave delivers personalized, trustworthy guidance without crossing into clinical decision-making?
Dave is designed to support, not replace, healthcare professionals. It personalizes guidance based on each patient’s cancer type, treatment stage, and emotional needs, yet explicitly avoids diagnosis or treatment recommendations. To ensure this boundary is never crossed, we’ve implemented multiple safeguards: a curated knowledge base, strict response filters, human-in-the-loop oversight, and continuous audits by clinical and regulatory experts. Powered by our proprietary Arch and Braincell technologies, Dave connects users with trusted resources and facilitates informed conversations with their care teams, ensuring trustworthiness, safety, and accuracy at every step.
2. What surprising patterns or unmet needs emerged from your dataset, and how did they shape Dave?
We discovered that emotional support, practical advice, and peer experiences are just as vital to patients as clinical information. Many users sought help navigating anxiety, daily life, and treatment challenges alongside their medical journey. Dave responds to each patient’s unique needs by providing holistic, whole-person support, offering personalized guidance and connecting them to relevant community resources.
3. What does it mean for an AI to have emotional intelligence, and how do you test it in Dave?
Emotional intelligence means Dave is built to understand tone, context, and user emotion, responding with appropriate empathy, warmth, and support. This is especially critical in oncology, where emotional nuance can deeply impact patient experience. We rigorously tested Dave’s emotional intelligence through real-world conversations, expert reviews, and continuous user feedback loops.
In addition to evaluating response accuracy, we assess tone, language sensitivity, and user sentiment to ensure interactions feel human and compassionate. Because healthcare demands higher safety and trust standards, we’ve layered Dave’s general-purpose AI core with proprietary, domain-specific technologies that safeguard emotional alignment while preventing missteps, striking the balance between responsiveness and reliability.
