Vision of Avinasi Labs

We have laid out Avinasi Labs’ Short-Term and Long-Term Vision. This document is a must-read for all team members to ensure alignment, as well as for anyone who wants to understand what longevity is a

Author: Winnie Qiu, Co-Founder of Avinasi Labs

Avinasi Labs: Short & Long-Term Vision

Avinasi Labs is building a liquid foundation model for human aging. My co-founder, Dr. Albert Ying, and I started Avinasi Labs with a grand vision:

To solve aging, and, when we solve it, to make sure everyone in the world has equal access to living healthier and longer lives.

We came to this conviction from very different places. For me, longevity is inseparable from grief. I built my first company in pain after watching my father, my grandfather, and thousands of cancer patients face death. To me, building in longevity means reducing that suffering. For Albert, a physicist turned biologist, dedicating himself to longevity is about saving brilliant minds: people with only a short window left to explore ideas that might change the world forever.

We also quickly realized that we are not alone in this pursuit. In 2024, longevity companies collectively raised over $8 billion, with some of the most prominent names including Retro Biosciences funded by Sam Altman, Altos Labs by Jeff Bezos, Unity Biotechnology by Peter Thiel, and NewLimit by Brian Armstrong. The momentum is undeniable. Yet beneath the surface, we see layers upon layers of unresolved problems:

• Scientific limitation – we still don’t fully understand the causal mechanisms of why we age.

• Linearity – biology and clinical development remain linear, limiting the speed of interventions and drugs (traditional trials simply do not scale for longevity).

• Infrastructure gap – with the exception of Singapore, global healthcare systems remain focused on “sick care” rather than preventative care, with incentives and regulations designed to treat disease instead of preserving health and longevity.

• Equitable longevity – access to both health and longevity remains deeply unequal worldwide.

Each of these challenges is monumental on its own, but they are also intertwined. And behind each lies another layer of causality. For example, our inability to pin down the drivers of aging is not only a scientific challenge; it reflects how data is siloed in labs, inaccessible to others, and generated unevenly across populations. There are few incentives for data owners to share. There is also a scarcity of high-quality, standardized data, partly because of affordability and technological barriers.

I once had a public debate with my friend Dr. Andrew Critch, the co-founder of HealthcareAgents.com, about whether we already have enough data to make life-saving decisions today. We agreed that the bottleneck is not the absence of data but access to it—and, equally, the need for new, unique datasets that illuminate what we don’t yet understand. This is why research such as Vittorio Sebastiano’s work on how women age, or João Pedro de Magalhães’s studies of long-lived animals, just to name a few in the longevity field, is so important. Yet even when data is produced, translating it into biomarkers, therapeutics, or interventions remains costly.

Yes, the cost of sequencing a whole genome has plummeted, from around $1 million in 2007, and about $100,000 when Steve Jobs had his genome sequenced, to under $200 today. But vast areas of biology are still underexplored. Epigenetic profiling and proteomics remain orders of magnitude more expensive than genome sequencing, and these layers of biology may hold some of the deepest causal insights into aging.

As we unravel these issues, the list of problems only grows longer. What Albert and I see as most powerful, and what could touch every part of this picture, is solving the data bottleneck itself. That has become our short-term vision:

To build a liquid, open-source foundation model for human aging, one that can be used by both the general public and institutions, self-fueled by participation, and capable of generating useful insights across 8 aging-related categories and at least 100 targetable assets within the next two years.

This vision makes sense because it allows us to focus on what we do best—finding causal signals in aging (Albert’s work on MethylGPT, the causality-enriched methylation clock, and my previous startup experiences in patient data, Web3 and hardware), while at the same time addressing the structural barriers around infrastructure, incentives, and translation. It also creates genuine upside for both individuals and institutions. Donation models or top-down mandates rarely move the needle; people contribute when they see clear value. Think of OpenAI: millions of people willingly pour their data, questions, and personal stories into ChatGPT every day, because for $20 a month they receive something transformative in return. The lesson is simple: if we build a system that delivers outsized value for the smallest effort, people will use it, and in using it they will strengthen it.

The same logic applies not only to researchers but to everyday users. Millions of people already check their Apple Watches, Oura Rings, or Whoops every morning, using the limited insights these devices provide to decide whether to train harder, sleep longer, or adjust their routines. But those insights are shallow, because they come from closed, proprietary systems that cannot connect to broader streams of biological data.

And yet even within the confines of closed datasets, we have seen what is possible. AlphaFold stunned the world by predicting protein structures at scale, built on relatively limited data. Chai Discovery has raised antibody hit rates from fractions of a percent to the double digits, and XtalPi has turned once “undruggable” targets into tractable ones. These are extraordinary achievements made within closed walls. Now imagine what becomes possible when data flows freely, when experiments are no longer bound by human labor, time zones, or geography. A liquid, open network could let AI agents run experiments continuously, testing and validating hypotheses around the clock, generating discoveries that are immediately shared and built upon, instead of locked away.

We are convinced that once data flows freely, liquid, open, validated by networks of agents and participants, the pace of discovery will accelerate beyond what humans can manage alone, while still producing research of deep value for humanity, from feeling energized all day long to living a fulfiling 150 years or longer, from healthspan to lifespan, from less suffer to more time for exploring possibility of life.

Our short-term vision of building a liquid, open-source foundation model for human aging is inseparable from our long-term vision of solving aging for everyone. We hope more people will join us.

Next, we will go deeper into how we are bringing this vision to life.