AppliedAI’s Opus makes compliant and reliable agentic automation available to every company – News

Abu Dhabi-based AppliedAI, which prides itself on being “the world’s most boring AI company,” today announced the global availability of its compliant agentic business process management tool Opus, enabling enterprises in the most heavily regulated industries to embrace AI automation.

Founded in 2021, AppliedAI specializes in automating mission-critical business workloads for highly regulated enterprises in sectors such as healthcare, financial services and government. It does this through the Opus platform, which is based on a specialized “Large Work Model” and a proprietary Work Knowledge Graph to deliver supervised automation outcomes that are both transparent and fully auditable.

With Opus, the Large Work Model generates business workflows as structured task sequences that can integrate human reviews for oversight, while the Work Knowledge Graph helps to address the complexity and precision needed for unambiguous workflow generation. It encodes industry-specific procedural and operational knowledge into graph-based representations, ensuring that the workflows are highly accurate and compliant with domain-specific regulations.

In layman’s terms, Opus is built to handle the mountains of paperwork that companies in financial services, insurance, healthcare and so on normally have to process manually. For instance, reviewing complex medical billing records and insurance claims. Enterprises often will hire teams of human workers to perform these tasks, or else outsource them to specialist companies such as Cognizant Inc. or Wipro Inc. — either way, the work takes hours and can cost millions of dollars in employee wages.

AppliedAI founder and Chief Executive Arya Bolurfrushan says Opus can do this work faster and more affordably, automating most of it while plugging human reviewers into the process to ensure everything is done right. “All our work is what we call high cost of error workflows where often human lives are on the line,” Bolurfrushan told Forbes in a February interview. “You can’t afford hallucinations and that’s why we have an essential harmony between our AI and human agents.”

Self-serve AI automation

Until now, AppliedAI has acted like more of an AI consultancy, with Opus only being made available to select customers and its integration with their systems closely supervised to ensure those deployments achieve the desired results. However, the company now believes that Opus is ready for prime time, and so it’s making it available as a self-serve platform for the first time, expanding its reach to any business that wants to realize the benefits of AI automation.

AppliedAI believes there will be big demand for a platform like Opus. It cites the Massachusetts Institute of Technology’s 2025 State of AI in Business report, which found that even now, 95% of generative AI pilot projects ultimately fail, costing businesses millions of dollars in wasted research costs. According to Bolurfrushan, the 5% of AI pilots that succeed do so because they’re based on systems that can learn and adapt, integrate seamlessly into business workflows, and be held accountable.

That’s exactly what Opus is designed to do. It sets out to identify, assess and prioritize opportunities for AI automation, then generates workflows that are fully compliant from the start, which can integrate with any platform, and learn as they progress. It has some impressive credentials, too. AppliedAI built the tool in collaboration with Palantir Technologies Inc., McKinsey & Co., Metalab and Group 42 Holding Ltd., better known as G42, specifically to help enterprises move from AI experimentation to production.

“The world doesn’t need more pilots. It needs production,” Bolurfrushan said. “Pilots fail because they’re static — don’t learn, don’t integrate, don’t deliver ROI. Opus is the antidote: self-evolving workflows that empower process owners, collapse design and iteration time, embed compliance and human oversight that compounds durable value with every use.”

Holger Mueller of Constellation Research Inc. believes that AppliedAI can be a disruptive force in enterprise AI, not only because of the Opus technology, but also because of its location in the Middle East.

“AI doesn’t just promise to transform software automation, but also recalibrate where the innovation happens,” the analyst said. “The Middle East has already emerged as an important funding center for AI, and now it has a startup of its own with AppliedAI coming out of Abu Dhabi. This can be a factor for sovereign automation, with all of its IP coming from the Middle East, it may well make the company a more agreeable AI partner for local companies.”

Opus is being made available through the Opus Cloud platform, which combines enterprise compliance and reliability with consumer-grade usability, so that companies of all sizes can get up and running with agentic business process automation in the fastest time possible. For larger enterprises, Opus Enterprise offers dedicated support and extended governance capabilities.

Bolurfrushan says Opus has already delivered tremendous benefits to customers in multiple industries. For instance, he claims that it has reduced energy procurement cycle times from an average of four hours to just 10 minutes, while in manufacturing vendor onboarding, it can reduce vendor processing times from two weeks to as little as five minutes. These are just two of many examples he provided.

“We are not in the business of hype,” Bolurfrushan added. “We are in the business of outcomes. Opus is designed to empower process owners to rewire their operations and deliver measurable ROI with discipline and oversight.”

It’s not just Bolurfrushan who is singing the praises of Opus. McKinsey analyst Tamim Saleh said what sets it apart is the level of trust, compliance and operating discipline it provides. “It compresses months of process re-engineering into minutes, pairs agentic AI with human judgment, and enables responsible scale in industries where failure is not an option,” Saleh said. “This is a credible path to rewiring operations.”

Image: AppliedAI