OrbitronAI unveils regional models for automated business decisions

OrbitronAI has introduced System 1 Models, an artificial intelligence service designed to automate business decisions while keeping customer data within Saudi Arabia or the United Arab Emirates.
The enterprise technology company says the service processes information on infrastructure hosted in the two countries and returns structured, probability-based assessments through a single application programming interface, or API. The offering targets organisations seeking automated decisions without sending sensitive information to processing facilities outside their chosen jurisdiction.
The announcement covers classification, ranking, risk assessment and intent detection. Rather than generating conversational responses, the models evaluate supplied information against predefined possible answers, allowing applications to use the resulting probabilities or scores in operational workflows.
OrbitronAI says requests are processed within the selected hosting region and that data does not leave the relevant national boundaries. Its public product information identifies separate Saudi and UAE hosting options, although detailed technical documentation on data retention, independent security testing and service performance has not been published alongside the announcement.
Access is initially being organised through a waiting list. Interested developers and organisations can specify their preferred hosting region and expected monthly volume of API calls. The company has not publicly disclosed pricing, a general availability date or confirmed external customer deployments for the new service.
Saul Adomaitis, OrbitronAI’s co-founder and chief executive, said the development was intended to connect AI outputs more directly with decisions taken inside business processes. He identified fast, actionable responses and regional data sovereignty as important requirements for enterprise adoption.
Ashu Gupta, co-founder and chief AI officer, said the company had already used System 1 Models within its own infrastructure and products before deciding to make the capability available to other developers. His comments indicate an extension of an internal technology rather than the creation of an entirely separate enterprise platform.
The models support several types of responses. A binary assessment returns probabilities for two alternatives, while a multiple-choice assessment assigns probabilities to available options. Another format produces a score along an ordered scale, potentially allowing software to distinguish between different levels of operational risk.
An application could, for example, submit information about a transaction or workflow and ask whether it requires escalation. The service would return an assessment based on the permitted answers, leaving the integrating application to determine what action follows. Such examples illustrate the proposed workflow rather than establish measured accuracy in a live deployment.
The distinction matters because a probability is not itself a verified outcome. Businesses using automated assessments still need to decide acceptable error rates, review thresholds and the circumstances in which human approval remains necessary. OrbitronAI has not published independent benchmarks demonstrating how its models perform across particular industries or languages.
The company describes its approach as suitable for applications requiring prompt decisions rather than extended conversations with an AI assistant. It also positions the service for developers who want predictable response formats that can be incorporated into existing software without extracting an answer from free-form generated text.
Data residency is a separate consideration from model accuracy. Hosting a service inside national borders can help organisations meet internal information-handling policies and certain contractual requirements, but compliance depends on the applicable rules, the information processed and the full technical arrangements. A provider’s residency claim alone does not establish regulatory approval for every intended use.
OrbitronAI also develops enterprise AI agents through its NovaOS platform, which it describes as a system for coordinating workflows, connecting enterprise applications and maintaining governance controls. The new API offering concentrates on individual structured decisions rather than the broader orchestration of multiple agents.
The company has an industrial AI partnership with Aramco Digital aimed at developing applications for capital projects, supply chains and industrial operations. That agreement provides context for OrbitronAI’s enterprise focus, although neither company has identified System 1 Models as a component of a specific joint deployment.
The public description does not identify the underlying model architectures, the precise infrastructure operators or contractual service-level commitments. Those details would be relevant to procurement teams assessing operational resilience and integration requirements.
On its product page, OrbitronAI illustrates the service with sample inputs and probability scores, and notes that the final API request format may change. The examples are explicitly illustrative rather than evidence of production performance.