A future of human health breakthroughs powered by clinical data.
We're building the technology infrastructure to make it possible.
Our Vision
Lighten envisions a future where clinical data drives the next generation of human health breakthroughs — from uncovering personalized treatment insights, to accelerating drug discovery, to driving adoption of innovative therapies.
To get there, healthcare needs a new kind of ecosystem in data, infrastructure, and AI — one that brings quality, transparency, and clinical rigor to the forefront, at a scale that was previously unthinkable.
That's what we're building.
Where Clinical Expertise Leads the Product
Most AI companies in healthcare are built by technologists alone, with clinical experts relegated to labeling data for their models. We took a radically different approach.
At Lighten, clinicians sit at the center of the product — shaping how it reasons, how it handles ambiguity, and how it reflects the realities of clinical practice.
We embrace the nuance of clinical data rather than silently ignoring it, unlike what most AI systems do.
Our Leadership Team
Xinkun brings deep expertise in AI and causal inference research, combined with a track record of designing and building technology and systems innovations from the ground up. While at Stanford, Xinkun developed state-of-the-art machine learning methodology for personalized treatment effect estimation—work that was rapidly adopted by leading Silicon Valley companies to power decision-making at scale. She went on to Actuate Innovation, a DARPA-inspired R&D organization, where she led the design of ambitious technology systems tackling problems too hard for industry or academia alone.
Xinkun founded Lighten with the mission of accelerating healthcare innovation and advancing precision health through breakthroughs in data, infrastructure, and AI.
Lighten's mission is deeply personal. Xinkun started Lighten in the third trimester of her first pregnancy. She's grateful to be working on something that carries deep meaning—something she hopes will make the next generation healthier. When not at work, Xinkun spends all her free time with her two young children.
Xinkun holds a PhD in Computer Science from Stanford University, and a Bachelor of Science in Electrical Engineering and Computer Science from MIT.
Kathryn brings over a decade of clinical and healthcare data experience to her role as Head of Product and Clinical Strategy. She was most recently Head of Clinical Abstraction at Verana Health, driving clinical strategy for curating ophthalmology, urology, and neurology unstructured EHR notes. She also led abstraction teams across cardiology and oncology and developed the clinical validation plan for an FDA-approved predictive ML medical device at Tempus.
Previously, Kathryn supported the NSQIP national outcomes registry at the American College of Surgeons, and began her career as an ICU and ED nurse. She has deep expertise in EHR data abstraction and building clinical data teams to power national registries, regulatory studies, and AI model development across therapeutic areas.
When not at work, Kathryn enjoys golfing, gardening, teaching her border collie, Rue, new tricks, and spending time with her two boys.
Pierre brings over a decade of applied AI and venture experience to his role as Head of Engineering. Most recently, he was Co-Founder and CTO of Stitchflow, an Index Ventures–backed enterprise SaaS company, and previously served as Staff ML Engineer at Okta, shipping real-time ML services to enterprise customers at scale.
Earlier in his career, Pierre built production NLP pipelines for healthcare and enterprise applications at Snorkel AI. He also co-founded Woebot with Andrew Ng and Alison Darcy — the pioneering AI-based mental health chatbot.
When not at work, Pierre enjoys MMA and snowboarding, and spending time with his son.
Proudly backed by
What we stand for
There is no "good enough"
Anything less than the highest quality standards means the wrong conclusion about a real patient.
Innovate tirelessly
We continuously out-innovate ourselves and redefine what's possible.
Responsible AI is AI built with humility.
Because silent failures are unacceptable when decisions shape patient care.