The Egyptian Refining Company (ERC) is conducting a trial run of an AI agent designed to predict equipment failures before they occur, tackling a high‑stakes problem that can cost millions per incident, according to Ahmed Heikal, Chairman of ERC parent Qalaa Holdings.

“We don’t have the luxury of doing it wrong,” said Heikal during the forum, organized by online news outlet Enterprise. He noted that each day the refinery sits idle costs $5 million in lost revenue, and with a seven‑day restart, a single outage means a $35 million hit.

ERC’s AI system tracks pressure, temperature, vibration, and flow data in real time to spot early warning signs. It learns the difference between healthy and faulty machinery patterns, and when risks appear, it automatically compiles a maintenance report with past work orders, root‑cause analyses, and spare‑parts checks. Engineers then review the findings and schedule repairs before a breakdown can halt operations.

Heikal also shed light on other AI agents used by other Qalaa subsidiaries. However, he cautioned that  AI usage comes with a downside. Describing cybersecurity as the biggest and most immediate risk associated with technology, Heikal noted that three oil and gas companies recently paid “large sums” to decrypt data after being hit by cyberattacks.

The ERC refinery is a $4.3 billion greenfield project designed to convert lower-value fuel oil into higher-value middle and light distillates, including diesel, LPG, and jet fuel, for the Egyptian market. The refinery forms a key part of Qalaa Holdings’ energy portfolio.

Qalaa Holdings is an investment company with operations spanning energy, mining, cement, agriculture foods, transportation and logistics, and printing and packaging.