I was born and raised in sunny Portugal and currently live in Porto (which isn’t always so sunny 🌧️). I enjoy working at the intersection of Machine Learning, Research, Product, and Engineering, building products that meaningfully impact users' lives.
I currently lead Applied AI at Sword Health where we’re reinventing how patients access and receive care—making treatment more human, clinically effective, and scalable. My focus is on building and scaling machine learning systems that drive Sword’s mission of freeing 2 million people from pain.
When I joined Sword Health, the entire company could fit into a single office room in downtown Porto. Today, we’ve grown to over 1k with offices in Porto, Lisbon, and NYC.
Some of my (and my team’s) contributions so far, in reverse chronological order, include:
- Building Chapter 2 of Phoenix, a proactive, continuous, deeply personal AI care companion available 24/7, built around natural conversation, long-term clinical memory, and clinician oversight.
- Creating Arbor, a framework for reliable navigation of critical healthcare conversation flows (read the technical report here).
- Organizing Sword AI Summit 2025.
- Post-training (SFT/RLHF/RLAIF) of LLMs on health data curated by clinicians, optimized to maximize clinical outcomes.
- Building Gondola, a human feedback interface powering continuous evaluation cycles.
- Designing and launching 10+ AI features behind the AI Feed, Sword's core care-management system supporting clinical decision-making, engagement, and administrative workflows at scale.
- Building Phoenix, a clinically safe, context-aware, and engaging voice agent (built by combining STT, LLM and TTS models) that guides therapy sessions and sustains member engagement.
- Creating the AI Care Coordinator, which resolves incoming member support tickets across email, chat, and SMS
- Growing the Applied AI team from 3 to 20+ team members, leading 4 teams
- Co-organizing and presenting at the Sword AI Summit 2024, with ~700 attendees (pulled together in under two months 😅)
- Co-inventing 10+ patents
- Building human motion tracking and analysis models that have already powered 7M+ AI Care Sessions
- …and other random things you’d expect (and not expect) in a fast-growing startup 🚀
Although building machine learning products in healthcare at scale is cool nothing tops receiving feedback like this 🥹.
I studied Biomedical Engineering at the Faculty of Engineering, University of Porto (FEUP). For my master’s thesis, I developed a computer vision model for human motion tracking, well before off-the-shelf models became a commodity.
When I’m not working you can find me traveling, reading, hiking, planning DIY projects that might not get started (or finished), exploring new restaurants, or saving my feet from the bites of my cat, Flecha.
📬 Feel free to reach me at anaclaraferreiramatos@gmail.com.