Iranian AI Powers Global Oil Industry Innovation

Pasargad Energy Development Company has launched a domestically developed platform for integrated well, reservoir, and surface facility management, incorporating Iran’s first upstream intelligent assistant (UIA) based on Large Language Models (LLMs), significantly improving access to technical knowledge and reducing search times.

According to the Public Relations Office of the Pasargad Energy Group, as reported by SHANA, in an era where rapid and precise access to specialized information is crucial for effective decision-making and competitiveness in strategic industries, Pasargad Energy Development Company has made a major leap in adopting cutting-edge technologies. For the first time in Iran, the company has deployed artificial intelligence in its localized integrated platform for managing well, reservoir, and surface facilities, operational since January 2025. This platform is now being used in both production and drilling operations.

Furthermore, the company has introduced the first upstream intelligent assistant (UIA) for the oil industry, leveraging LLMs trained on high-quality international data sources. This system provides targeted, streamlined access to domain expertise, enhancing the accuracy of responses and dramatically reducing the time needed to retrieve technical knowledge.

This initiative, regarded as one of the few of its kind globally, marks a significant step toward the digital transformation of Iran’s oil industry knowledge infrastructure. As the global and domestic oil sectors increasingly pivot toward artificial intelligence, the Ministry of Petroleum’s Head of Media and Publications, Parastoo Younesi, along with a delegation from SHANA, visited the WRFM (Well, Reservoir, and Facility Management) Center at Pasargad Energy.

During the visit, Saeed Dehghani, CEO of Sepehr Pasargad Oil and Gas Operating Company, emphasized that since February 2024, the development and deployment of business intelligence and AI systems in oil and gas field development and production have been prioritized by the Pasargad Energy Group. “As you know, exploration and production companies typically do not manage wells, reservoirs, and surface facilities in an integrated, real-time manner,” Dehghani said. “In January 2025, the WRFM Pasargad platform was officially launched, and from that very day, full-scale AI integration was set as a target to be achieved within a year.”

Thanks to the tireless efforts of the company’s specialists and the collaboration of top academic institutions and knowledge-based companies, artificial intelligence capabilities were added to the platform’s production and drilling functionalities in less than five months. “Given the current circumstances, we cannot acquire such platforms from foreign vendors, and even if procurement were possible, deployment and operation would be fraught with challenges,” Dehghani added.

The achievement underscores Iran’s growing self-sufficiency in high-tech applications within the oil and gas sector and positions Pasargad Energy as a regional leader in AI-powered upstream solutions.

He further emphasized that beyond real-time data acquisition, the WRFM Pasargad platform also significantly facilitates data analysis and interpretation. “This platform not only enables standardized data collection and storage in centralized databases but also supports real-time data analysis. It allows users to detect patterns and trends, trigger necessary alerts, and visualize all information within an integrated system,” he explained.

As a result, both raw and interpreted data are made instantly available to decision-makers in a unified environment, greatly accelerating the decision-making process.

The platform encompasses a wide range of management systems, including real-time well production monitoring, pressure and temperature alert configuration systems for wellhead and oil collection/transfer units, real-time drilling operations management with performance-to-plan comparisons, surface facilities maintenance systems, well production forecasting, and Rate of Penetration (ROP) prediction—comprising around 12 business intelligence modules and 3 artificial intelligence modules. These capabilities are currently being utilized in the development and production operations at Sepehr and Jofeir oil fields.

Dehghani noted that one practical implementation of the platform has been at the Sepehr and Jofeir oil fields. “In this project, we achieved system uptimes as high as 98%, with only 2% production downtime. Moreover, this platform has had a major impact on time management and cost reduction,” he said. For example, in one well, leveraging the capabilities of this platform prevented production halts and increased output—generating an estimated added value of $150 million.

He went on to explain that one of the key cost drivers in reservoirs with asphaltene challenges is the lack of real-time monitoring and analysis of well pressure and temperature. If the fluid production rate and dynamics create conditions conducive to asphaltene deposition in the wellbore or surface facilities, this can lead to production shutdowns, costly well interventions (rigless or rig-based), or require the continuous injection of chemical inhibitors—all of which impose significant financial burdens.

“In companies with a history of using asphaltene inhibitors, costs have typically ranged from $0.70 to $1.40 per barrel. Through optimized well management using the WRFM platform at Pasargad, we’ve aimed to either eliminate the need for chemical injection altogether or reduce it to a minimal level. As a result, not only have the wells remained online, but the operating expenses (OPEX) for Fahlyani wells have been significantly reduced compared to other fields,” he concluded.

The CEO of Sepehr Pasargad Oil and Gas Operating Company emphasized that this platform enables a true shift toward full-field digitalization—what is known as a Digital Oil Field. “While SCADA systems have been installed in some oil fields across Iran, the absence of an integrated platform for data analysis across wells, reservoirs, and surface facilities has prevented full maturity,” he noted. “We had sensors and measurement systems in place, but no unified platform that could leverage real-time data for strategic decision-making. These systems were mainly limited to production operations and did not cover drilling or the navigation of horizontal wells.”

Dehghani highlighted that the achievements at Pasargad—especially with the support of the Group’s CEO—represent a true step toward building a digital oil field. “We didn’t just adopt a platform; we custom-designed it to match our specific needs. The platform also receives real-time data from drilling rigs and downhole operations. It’s not merely a monitoring center—it includes decision-support capabilities, predictive alerts, and AI-powered trend forecasting.”

Moreover, since the platform allows users to upload production forecast plans, drilling schedules, and geological models, it provides a major advantage by enabling comparisons between actual operations and predefined plans. “Specialists can even set custom thresholds for pressure, temperature, weight, speed, and more—so if any values exceed defined limits, the system will automatically trigger alerts. It operates 24/7 without the need for constant human supervision.”

Dehghani stated that in 2024, the business intelligence layer of this platform was completed according to its designated mission. “So far, we’ve achieved three major milestones. The first is intelligent production forecasting.

Instead of relying on time-consuming 3D reservoir simulation models, we can now predict a well’s production performance in under 10 seconds using artificial intelligence—helping us quickly assess whether the well is operating at optimal efficiency.”

He added, “We are also working toward enabling the AI to detect the root causes of well performance issues. The vision is to eventually reach a point where the system itself recommends stabilization and enhancement strategies for production.”

Another module where significant progress has been made is drilling rate optimization. “In drilling operations, achieving optimal Rate of Penetration (ROP) is a critical objective,” Dehghani said. “Our next goal is to roll out intelligent prediction of drilling-related issues—an essential concern in the oil industry—which we expect to launch within the next 3 to 4 months.”

This smart prediction will utilize two approaches: analysis of surrounding well data and real-time drilling data. In both cases, artificial intelligence will identify and report problems in a fraction of a second.

“Through our efforts, we realized the pressing need for students and professionals in the oil and gas sector to have access to a variety of high-quality knowledge sources,” he added. “To reduce the time required for accessing this information, we needed a platform similar to ChatGPT. However, platforms like ChatGPT have a fundamental issue—they provide vast breadth but limited depth. You’re essentially dealing with an ocean of shallow information that’s not specific or detailed enough for specialized oil industry needs.”

As a result, Dehghani said, “we first acquired the knowledge needed to build a foundational platform tailored for our industry. Then we moved forward with the development of our own specialized system. By leveraging the capabilities of advanced Large Language Models (LLMs)—which have recently taken the world by storm—we created a platform that truly serves the technical needs of oil and gas experts.”