Governments are becoming major buyers of AI, and most AI startups have no idea how to sell to them. The tools that exist were built for US contractors. Bonaparte finds where a startup fits in public demand, in any country, and maps the path to the buyer.
Countries want control over their compute, data and models. That creates a new class of buyer: ministries, agencies and state-backed funds procuring AI as strategic capacity, not as IT.
Source: Grand View Research, Sovereign AI Market Report (2026).
Selling to a state is a different sport from selling to an enterprise. Founders who've mastered the second usually find the first opaque.
National strategies, budget lines and political announcements signal what a government will buy. By the time a tender is published, the shape of the winner is often already set.
Procurement rules, buying entities and partnership models change from Washington to Brussels to Abu Dhabi. No startup can learn them all at once.
Knowing that a ministry needs something is half the work. Knowing who sponsors it, who influences it and who can make the introduction is the other half.
The category is validated: recent rounds show that helping companies win public contracts is a venture-scale business. Most of these tools are built around the US procurement system and the companies that already know it.
Four stages, one run. The founder stays in control: nothing goes out without their review.
From a website, Bonaparte models what the company sells and which public missions it could serve.
Screens active RFPs, open procurement, announced programmes, planned demand, political signals and strategic gaps, then ranks them against the offer.
Builds the influence network around the buyer: who decides, who sponsors, and who can make the introduction.
Proposes who to contact, the goal of the conversation and a tailored first message, ready for the founder's review.
An AI company doesn't win a government with a better model alone. It wins by meeting several contracts at once, and only one of them is technical.
Théo's master's thesis at Bocconi University studies how frontier AI model companies build sovereign partnerships with governments, through comparative case studies of OpenAI in the UAE and Anthropic in Saudi Arabia. Bonaparte turns that framework into software any AI startup can use.
Bonaparte sits inside Sovereign AI SF, which also runs research and curated salons. That gives the product a built-in channel to the founders who need it and the public-sector people they need to reach.
Small discussions among AI founders, government GTM leaders, policymakers, public-sector operators and investors. Salon #01, "The Sovereign AI Stack," is planned for November 2026.
Practical analyses on market entry, procurement, sovereign partnerships and turning pilots into deployments. These build credibility and pull in the right founders.
A public view of where governments are signaling AI demand. It serves as a top-of-funnel tool that leads into the agent.
No revenue or customers yet. The milestones below are the evidence we're setting out to build.
AI startups running Bonaparte on real government pursuits, with weekly feedback on the quality of what it surfaces.
One introduction made through Bonaparte that turns into a real buyer conversation, then a pilot.
Which markets (US, EU member states, the Gulf) give AI startups the shortest route to a first contract.
Whether value is best captured by subscription, by success on contracts won, or by a mix of both.
We're opening conversations with angels and pre-seed funds who can help AI startups win their first government contracts.
Connect with Sovereign AI SF → See the product