Abu Dhabi: ADNOC and AIQ have successfully completed the proof-of-concept trial of ENERGYai, the world’s first agentic artificial intelligence (AI) solution tailored for the energy sector.
This AI solution integrates a 70-billion-parameter (LLM) with over 50 years of ADNOC’s expertise and petabytes of proprietary data to optimize operations and boost efficiency across the company’s energy value chain.
The 90-day trial demonstrated that agentic AI—AI ‘agents’ trained in specific tasks—can significantly enhance upstream exploration.
ENERGYai
ENERGYai delivered rapid, precise, and detailed seismic survey analysis, improving the pace and accuracy of exploration activities. It also provided actionable insights to optimize production at ADNOC’s existing wells. The AI system communicated these insights in intuitive natural language, allowing engineers to interact with the tool effectively.
Notable results from the trial include a 70% improvement in seismic interpretation accuracy and significant advancements in reservoir monitoring and anomaly detection.

Additionally, ENERGYai was able to enhance the quality of data, increasing reliability by detecting errors, standardizing formats, and enriching datasets for better usability.
Musabbeh Al Kaabi, ADNOC Upstream CEO, highlighted the promising results of the ENERGYai proof of concept, emphasizing its potential to drive value and sustainable energy production. Al Kaabi stated that, by leveraging ADNOC’s vast data, ENERGYai will empower the workforce and help future-proof the company, with the goal of becoming the world’s leading AI-enabled energy provider.
Magzhan Kenesbai, Acting Managing Director of AIQ, credited the success to the close collaboration between ADNOC’s subject matter experts and AIQ’s team of over 100 specialists. The robust AI infrastructure established during this phase forms the foundation for even greater achievements in subsequent phases.
Following the proof-of-concept trial, the first operational, scalable version of ENERGYai is set to be completed in H1 2025.
This version will include five fully operational AI agents focused on subsurface operations. It will be test-deployed across several upstream assets, with plans to scale its use to thousands of additional wells.

