Covid lockdowns had just lifted. On the back seat of a safari truck, bouncing along a pot-holed road in Pondoland, a long-time collaborator was telling me about his move to Singapore to lead the Duke-NUS Global Health Institute and the Centre for Outbreak Preparedness. He described how Next-Generation Sequencing was moving out of specialist labs into public health systems and clinics. The pandemic had pulled the future forward: sequencers becoming cheaper, more portable, and embedded in routine decision-making.
“But a gene sequence without context,” he said, “is like catching a raindrop in a storm, but not knowing the storm’s severity, whether it would make landfall, and how this will evolve”.
Context is everything
We need to understand not just what pathogen is there, but where, in whom, how fast it’s changing, and what it means in the real world. That requires combining genomic data with reported case loads, clinical outcomes, demographics, mobility, animal health, wastewater signals, climate and weather, health system capacity, supply chains, and more.
And we need to do this in ways that countries can trust: respecting data sovereignty, embedding clear governance, and using AI in ways that are explainable and accountable.
“That’s what we’ve been building IXO for” — it clicked!
That insight opened a new road to today.
Joyce Teo
We are building a co-operating system for infectious disease intelligence
Over the past decade with IXO, we have been working on building an Internet of Impacts: decentralised systems for turning complex, distributed data into verifiable evidence that can drive better decisions and better incentives. More recently, we have added Qi as the intelligence layer, which we describe as a human-AI intelligent co-operating system.
What clicked, was also an opportunity to reconnect with my medical and public health roots and apply everything we’d been building – secure data infrastructures, verifiable claims, digital twins, agentic oracles, and decentralised governance – to what we had all just experienced as one of the most urgent challenges of our time: infectious disease.
Over the next three years, I moved to Singapore, established a regional IXO office, and we began a formal research collaboration with Duke-NUS. In parallel, the Asia Pathogen Genomics Initiative (Asia PGI) was scaling up as a multi-country coordination and capacity development platform to accelerate the use of pathogen genomics for early detection and control of infectious diseases across Asia.
PathGen is the convergence point of these journeys.
IXO Spatial Web
What PathGen aims to achieve
PathGen is an infectious disease intelligence platform that integrates pathogen genomics with sovereign data and agentic AI to support faster, better outbreak decisions.
At its core, PathGen is designed to:
- Detect emerging threats earlier by combining genomic signals with clinical, environmental, mobility and population data.
- Shorten the time from “sequence to actionable insights” – for example, by augmenting decisions on treatment protocols, vaccine deployment, or vector control.
- Enable countries to cooperate on intelligence without giving up control of their data, through a federated, sovereign-by-design architecture.
- Serve as a digital public good for the region and beyond, built around interoperability, transparency, and shared benefit.
It is being developed as a collaborative effort, led by the Duke-NUS Centre for Outbreak Preparedness as the coordination hub and nerve-centre that connects more than 50 government and academic partners across around 15 countries.
The initial focus areas reflect real-world pressures:
- Tuberculosis, including drug-resistant forms, where genomics and contextual data can help detect clusters earlier and align treatment with resistance patterns.
- Arboviral diseases such as dengue, Zika and chikungunya, where linking virus trends with climate and mobility can improve risk forecasting and guide vaccination and vector control.
- Wastewater and environmental surveillance, where early signals can surface silent transmission before clinical cases surge.
In all of this, the aim is simple and hard at the same time:
shorten the distance between knowing and acting, without compromising trust.
IXO as the AI and data technology partner
IXO’s contribution to PathGen sits at the intersection of data architecture, data science, AI, and governance. From the outset, our role has been to help design and build PathGen as an infectious disease intelligence platform that is:
- sovereign-by-design
- decentralised and interoperable
- AI-powered, yet explainable and accountable
Concretely, that has meant focusing on a few key layers.
A sovereign, federated data architecture
PathGen is being built so that countries retain control over their raw data. Instead of centralising everything into a single warehouse, analytics are pushed as close to the data as possible. Only the insights that need to be shared are exchanged – in ways that can be governed, audited, and revoked.
IXO’s experience with decentralised data architectures and verifiable claims has shaped how we:
- Represent and enforce data access policies across jurisdictions.
- Use cryptographic proofs and verifiable credentials so that when a country shares an insight, others can trust what it represents without needing access to the raw data.
- Design the platform so that it can plug into diverse national systems, rather than demanding that everyone adopt a single stack.
Agentic AI for explainable decisions
PathGen is not just about more dashboards. It needs to help answer hard, time-sensitive questions:
- Is this new cluster indicative of a new variant or just better testing?
- Where should we prioritise limited sequencing capacity?
- When should we escalate from “watchful waiting” to a change in treatment or public health guidance?
Our work with agentic AI in IXO – building orchestrated AI agents as Agentic Oracles that can gather data, run models, and reason over complex contextual graphs – is being adapted to PathGen to support:
- Multi-source sense-making across genomic, clinical, environmental, and mobility data.
- Scenario exploration (for example, “What happens if vaccination is delayed by three months?”).
- Decision support that is transparent about assumptions, data sources, and uncertainties, rather than opaque “black box” predictions.
The goal is not to replace human judgement, but to give public health teams a more complete, timely picture so they can act earlier and with more confidence.
Governance, accountability, and benefit-sharing
Any regional intelligence platform must be anchored in clear rules and shared expectations.
IXO’s governance tooling and experience with verifiable impact systems are being applied to:
- Encode data access and sharing agreements into machine-readable policies.
- Provide clear audit trails of who accessed what, under which authorisations, and for what purpose.
- Support benefit-sharing arrangements so that contributions to the system – data, analytics, operational capacity – can be recognised and rewarded under agreed frameworks.
This is where “digital public good” moves from rhetoric to practice: countries are not simply data providers; they are co-owners of the platform’s rules and benefits.
Collaboration at the core
PathGen is a collaboration in the deepest sense. No single institution, technology provider, or funder could build this alone, and that is a feature, not a bug in the intelligent voluntary co-operating system that we are building.
A few elements of this ecosystem deserve to be named explicitly:
- Asia Pathogen Genomics Initiative (Asia PGI), as the multi-country regional collaboration network, supported by the Duke-NUS Centre for Outbreak Preparedness that convenes public health institutes, builds capacity, and integrates PathGen into a broader pathogen genomics roadmap for Asia.
- Catalytic funders – Temasek Foundation, the Bill & Melinda Gates Foundation, and Philanthropy Asia Alliance – whose support has enabled PathGen to be designed as a shared regional asset rather than a proprietary product.
- Technical collaborators – the University of Sydney, Sequentia Biotech, and AWS Health Omics – each bringing critical capabilities in bioinformatics, genomics, and cloud infrastructure to make a federated, AI-powered platform viable at scale.
- Participating country partners – including Indonesia, Malaysia, the Philippines, Thailand and Vietnam, among others in the Asia PGI network – who are not just “users” but co-creators, shaping priorities, governance, and deployment plans from the outset.
Why now – and what comes next
Asia faces a perfect storm with the convergence of: rapid urbanisation and population growth, climate disruptions, shifting patterns of mobility, and rising antimicrobial resistance. These are all amplifiers of infectious disease risk. At the same time, pathogen genomics capacity has grown rapidly, but is often under-utilised or siloed, especially in lower-resource settings.
Recent advances in AI make it possible to generate pathogen embeddings and to extract patterns from complex, multi-layered datasets in ways that were not feasible even a few years ago. The opportunity – and responsibility – is to use these tools to strengthen surveillance, response, and resilience, while building trust rather than eroding it.
From today’s public preview on 1 December 2025, PathGen now moves into an 18-month journey from proof-of-concept to a launch-ready platform, with pilots from early 2026 and a staged regional roll-out through 2027. During this period, we will:
- Co-create deployment plans with in-country partners, aligned to national priorities and regulatory frameworks.
- Harden the data, AI, and governance architectures to operate reliably in real-world conditions.
- Iterate on analytics and decision-support tools based on real use in TB, arboviral diseases, and wastewater surveillance.
- Continue to refine how sovereignty, equity, and benefit-sharing are encoded in the platform’s technical and institutional design.
A personal reflection
Sitting in that safari truck back in 2022, I couldn’t have predicted that the conversation was steering us down a road to launching a regional infectious disease intelligence platform, anchored in Singapore and co-created with partners across Asia.
The road ahead is now clearer than ever, even with its potholes and unchartered course.
For years, IXO has been working on how to turn complex, distributed data into trusted, verifiable intelligence for better decisions and better systems. PathGen is one of the most concrete, consequential applications of that work: bringing together genomics, contextual data, and AI to help countries see threats earlier and act faster, without giving up control.
From my perspective, this is what “AI for good” looks like when you take sovereignty, governance, and long-term public value seriously.
As we move from preview to pilots, I wonder if (when) we succeed, which potential outbreaks will quietly never happen because the signals were seen early enough, and the decisions were made in time.
At the cutting-edge of preventing, detecting, and responding to infectious diseases.
