People who readgovernments.
Bilby makes regulation, compliance and prediction software. Our mission is to use AI to improve how the world is governed. We started with the hardest case, China's government, and now read more than 40 countries.
From the desert to a map of who decides
How research on one government became software that reads more than 40.
Origins in the Australian desert
Bilby began in the Australian desert and grew into software that reads what governments do, at scale.
Research, put to work
Our founders brought years of research at Oxford and other universities into software that changes how organisations understand government.
The hardest problem first
We started with the hardest case in policy intelligence, China's government, and built Bilby to handle that complexity from the first day.
A map of who decides
We turn what governments do into a map of who decides, joining people, policies and outcomes across countries and years. Research analysts curate it.
Sovereign data and models
Every client's data stays private and secure. Our models are custom-built and sovereign, so clients keep full control of sensitive work.
Specialist knowledge, at scale
Decades of regional expertise are built into Bilby, so a team gets a specialist's reading of a market without hiring one for every market.
Data as our craft
The quality of our data is our signature. Every artifact we process is held to the same precision and structure, so you can act on it.
The team
Researchers who have followed Asia's governments for decades, analysts who curate what Bilby reads, and the engineers who build it.
Dr Ryan Manuel has more than 20 years of industry experience. A third-time founder, he has taught at Oxford, the Australian National University and the University of Hong Kong, and worked for BCG and the Australian government. He holds a doctorate from Oxford, where he was a Rhodes Scholar.
Satyadev runs the data infrastructure at Bilby. He has ten years of experience at startups and large tech companies, building data and machine learning products. Outside work he is an avid gamer and a cross-country motorcyclist.
David is a software engineer turned product manager, with more than five years of experience. He enjoys taking ideas all the way to something people rely on, and is driven by curiosity and careful execution.
Simon has lived and worked in Beijing, Hong Kong and elsewhere in Asia since the 1990s. A former Editor-in-Chief, Asia, for the Economist Intelligence Unit, he ran his own research and publishing business, Big Brains, for more than 20 years. He holds an MA (Hons) in politics from the University of Edinburgh.
A machine learning researcher and quantitative analyst, Samuel uses statistics, AI and large language models to find the signal in high-dimensional data. He holds a doctorate in statistical machine learning from the University of Oxford.
Damian works across data engineering, machine learning and AI, building scalable systems, data pipelines and predictive models in Python and SQL on the cloud.
Pallav is a software engineer specialising in AI agents, machine learning and software architecture, building systems that pair automation with thoughtful design. He has wide experience in data engineering and cloud infrastructure.
Harvey trains in-house models and analyses large datasets on the quant ML team. His speciality is machine learning for particle physics. Away from his desk he is likely cooking a steak with Tom Misch on the speakers.
Carlo is a software engineer who loves building sleek web apps and trying the latest tech. Outside work he is usually out with his camera, capturing life as it goes.
Jeremy works on both the research and the business side of Bilby. After work you will most likely find him with a craft beer in hand: he also runs a beer distribution business.
Marco has a background in international relations and enjoys working out how political systems function. Along the way he has collected a wide knowledge of the world, not all of it equally useful.
Nik supports our executives with the office and its administration, and loves travel and the outdoors.
In the press
Financial TimesHow data and AI make the patterns in government policy visible to business and investors.Read the article (opens in a new tab)
South China Morning PostA Hong Kong start-up reads government communications to see major regulatory change coming.Read the article (opens in a new tab)
Asian Business ReviewTurning government policy signals into insight, so businesses act before regulation shifts.Read the article (opens in a new tab)- Hong Kong BusinessMaking history predictive: a start-up helping firms get ahead of regulatory change.Read the article (opens in a new tab)
Working with Bilby
Who uses Bilby?
Governments, financial institutions and companies. Our particular strength is turning what governments do into data you can search and ask about.
How does Bilby fit with the tools we already use?
Bilby works in the browser. Its API and its MCP server let your own systems, and AI tools such as Claude and ChatGPT, ask it too; the Help Center's developer guides show how.
Can Bilby be shaped around our work?
Yes. Bilby is ready to use as it is, and we also build around your priority themes and widen the scope as you need.
How long does a project take?
Where the data is not yet collected, a project usually aims to be running in about five months, in phases: proof of concept, design and integration, testing, then rollout and handover.
How do I start?
Try Bilby for yourself, or write to team@bilby.ai and we will set up a briefing.
Put our teamto work.
Ask Bilby about any government it reads, or see how it works in the Help Center.
Try BilbyRead the Help Center