On March 27, 2023, shares of Presight AI began trading on the Abu Dhabi Securities Exchange under the ticker PRESIGHT. They opened at AED 3.52, against an IPO price of AED 1.34. By end of day, the stock had settled at AED 3.25: a 163 percent surge on first trading. The listing had attracted nearly AED 94.9 billion in bids, a 136-times oversubscription for a $496 million offering. Thomas Pramotedham, who had been leading the company for less than two years, was watching the market price a wager he had made in a different city-state about what sovereign clients are worth when a company builds for them properly.
The Map-Maker Who Stopped Working for One Government
Before Abu Dhabi, there was Singapore. Pramotedham spent years running Esri Singapore, the regional arm of the world's most widely deployed geospatial technology company, during the period when Singapore was assembling the data infrastructure for its Smart Nation program. His job was concrete: take geographic data, from population flows to urban density to emergency response routing, and turn it into instruments that ministries could use to govern by evidence rather than by inspection. Singapore's Urban Redevelopment Authority used the platform to analyze how residents, jobs, and transport related to specific plots of land. Government planners translated abstract city-planning scenarios into three-dimensional models precise enough to design, build, and develop with clarity. Pramotedham described spatial thinking as foundational to the program, and spent years making that argument inside the ministries where it counted.
The work built a conviction that ran counter to how most technology executives read the government client. In the standard technology sales narrative, governments are slow, risk-averse, and difficult: long procurement cycles, opaque decision-making, low margins. Pramotedham's experience in Singapore pointed in a different direction. A government client does not buy a product. It integrates a system into the operations of the state. Once that integration is established, the platform does not get reconsidered the way a SaaS tool gets reconsidered at renewal. It deepens: every additional dataset, every additional ministry that touches the system, every new use case the platform enables makes the relationship more embedded and more durable.
By 2020, G42, Abu Dhabi's flagship artificial intelligence holding company, had reached a version of the same conclusion from inside the emirate. G42 had spent three years building AI infrastructure at the scale of a city-state and had seen what sovereign data, health records, city operations, energy systems, could do when connected through intelligent systems. The group wanted a company that would take that model, productize it, and extend it beyond the UAE's borders. G42 founded Presight AI in 2020 for exactly that purpose.
In May 2021, they brought Pramotedham to Abu Dhabi to lead it. The brief was not to find product-market fit. It was to prove the sovereign intelligence model at scale, across multiple governments, before the window for first-mover advantage closed.
Seventeen Governments Before the Sales Pitch Was Obvious
The standard playbook for an applied AI company in 2021 was narrow: build one repeatable product, prove it in a single market, and scale horizontally. Presight took the opposite position.
Rather than narrowing to a single application layer, the company pushed deeper into the sovereign relationship itself: smart city infrastructure, public safety analytics, energy optimization, crisis response, healthcare management. Rather than anchoring to one jurisdiction, it moved toward governments that had the political will to transform and had not yet locked into contracts with entrenched vendors.
The wager was structural. A company that wins a government infrastructure deployment in Abu Dhabi does not have one client in the ordinary sense. It has a reference that other sovereign governments cannot easily dismiss. A company that builds the national AI architecture of Kazakhstan does not simply earn a contract. It earns the data relationship, the local engineering workforce, and the institutional trust that no competitor can replicate by arriving later with a more refined pitch.
Sovereign contracts carry real costs. Procurement cycles are long. Implementations are bespoke. Every deployment is an integration project, not a software installation. The margin for error on a major national infrastructure contract is not the same as the margin for error on an enterprise SaaS subscription. Presight's strategy accepted those terms because the ceiling was also different: the government that has embedded a platform into its public safety infrastructure, its energy grid, or its city operations does not reconsider it at the next budget cycle. It extends it.
The March 2023 IPO was the market's opinion on the wager. Nearly AED 94.9 billion in orders for an offering designed to raise $496 million. An oversubscription of 136 times. A debut-day surge of 163 percent on the Abu Dhabi Securities Exchange. Pramotedham told CNBC at the time that it was the right moment to capitalize on growth in the Middle Eastern market. The market agreed, and then priced a premium on top of that agreement.
"We see the MENA playing three critical roles: it will launch enterprise-grade AI, lead in sovereign data innovation, and promote ethical, applied AI." Thomas Pramotedham, Finance Middle East
Country by Country, One Sovereign at a Time
The geography of Presight's international expansion makes unusual reading. Burkina Faso, Côte d'Ivoire, Gabon, Kazakhstan, Albania, Montenegro, Jordan: the list of sovereign partners does not follow a population tier or a regional cluster. It follows a single criterion: governments with the political will to deploy national AI infrastructure and without an incumbent Western vendor already installed in the ministry.
G42's network opened conversations that most international technology firms could not easily enter. Pramotedham's team then had to turn those conversations into contracts by doing what most technology vendors avoid: the full integration work that sovereign clients actually require.
The Astana deal became the template. In a six-year, $190 million agreement with Kazakhstan's capital city, Presight committed to deploying an AI-powered smart city system to digitize urban infrastructure, manage traffic, and modernize public services. The contract structure required Kazakhstani companies to carry sixty percent of the project scope, and required Presight to establish an R&D laboratory and local engineering capacity inside the country. The knowledge transfer was not incidental. It was contractual, because governments that wanted to import AI infrastructure also wanted to own the capability over time.
The model compounded as designed. Kazakhstan's national government then appointed Presight to develop a national supercomputer and data center cluster. In July 2025, Kazakhstan officially launched that supercomputer. The smart city project in Astana had become the entry point for a national digital infrastructure program. One executed sovereign contract became two, then three, as each government relationship extended into adjacent mandates.
The financial results reflect this accumulation. Presight reported revenue of AED 2,213 million in 2024, a 24.3 percent increase year-on-year. In 2025, revenue crossed AED 3.03 billion, up 36.9 percent, with international markets growing 130 percent and accounting for 38.5 percent of total revenue, compared to 23 percent in 2024. The company had started the decade with international revenue representing less than five percent of its total. By the end of 2025, it represented more than a third.
On April 2, 2026, Presight signed memoranda of understanding with the governments of Burkina Faso, Côte d'Ivoire, and Gabon to build national AI infrastructure across West Africa. For Burkina Faso, the agreement included an AI Expert Factory to train local engineers and the establishment of the Ouaga Granit Valley Centre, a national hub for the country's emerging AI startup ecosystem. For Côte d'Ivoire, the partnership was positioned to establish the country as the leading hub for AI innovation across West Africa. The structure in every case was the same as Kazakhstan: infrastructure deployment with local capacity transfer built into the contract from the outset, not added as an afterthought.
The Government That Cannot Switch Is the Client Worth Building For
There is a version of the Presight story that reads as a lesson about access: G42 had the sovereign relationships, Abu Dhabi had the capital, and Pramotedham had the good fortune of arriving in a city-state that opened doors. That version is accurate as far as it goes. It misses the harder point.
The harder point is about the nature of the government client, read correctly. Governments are the worst customers by the metrics that enterprise software sales teams apply: long procurement, bespoke integrations, political risk, extended timelines. But they are the most durable customers by the metrics that compound over a decade: contract size, relationship depth, reference weight, and data accumulation.
Pramotedham had learned this in Singapore, where the geospatial data layer his team built was not a product the government could swap out at renewal. It was national infrastructure. The same dynamic operates at the AI layer, and at a larger scale. The government that has integrated its public safety operations, its city monitoring, or its national energy management with a single analytics platform is not reconsidering it at the next budget cycle. It is deepening it, extending it into adjacent systems, and training a local workforce on it.
The implication for founders building in applied AI or government technology is precise. The fastest path to durable global deployment is not the market with the most financially sophisticated buyers. It is the market with the highest political will to transform and the fewest incumbent vendors already installed. In 2020 and 2021, a dozen governments across Central Asia, West Africa, and Southeast Europe had that profile. They had the mandate and the urgency. Presight arrived first, did the integration work, transferred the capability, and built the kind of contract structure that compounds over years rather than quarters.
Pramotedham said in 2025 that the MENA region was positioned to set the pace on applied AI, with a focus on measurable outcomes while much of the world was still deciding which direction to take. By May 2026, with revenue surpassing AED 3 billion and international markets at 38.5 percent of total, the outcomes were measurable. No foreign AI firm built that sovereign client base through a sales trip. It was assembled, one government relationship at a time, by a company willing to do what the relationship required: embed deeply, transfer capability, and stay.