Technology
Aveva
Industry
Oil and Gas
Client
Daleel
Country
Oman
Goals
Develop an AI-powered operational intelligence chatbot within the existing DALEEL Portal. Enable users to interact with approved operational data and engineering knowledge through natural-language questions, supporting production surveillance, troubleshooting, analysis, and operational decision-making.
Challenges
Bringing together diverse operational and historical data sources, engineering references, and knowledge repositories while maintaining controlled, secure, and auditable access. The solution also needs to operate within DALEEL’s approved on-premises, CPU-based infrastructure and align with the existing portal experience.
Solutions
EMI’s scope includes requirements gathering and solution design, on-premises platform and local LLM setup, chatbot backend and UI development, integration with approved operational data sources and AVEVA PI System, and implementation of RAG, AI orchestration, analytics, and reporting using controlled backend data access.
Results
The solution aims to provide a secure, context-aware conversational interface to approved operational data, enabling users to monitor well and facility performance, detect deviations, review historical events, and access engineering knowledge. Deployment status, measured benefits, and user adoption must be confirmed before presenting these as achieved results.
|
DALEEL Generative AI Operational Intelligence ChatbotEMI was awarded a project to develop a Generative AI Operational Intelligence Chatbot within DALEEL’s existing portal. The initiative is designed to make approved operational data and engineering knowledge more accessible through natural-language interaction, supporting users across wells, facilities, production performance, deferment, and surveillance activities. The scope begins with stakeholder workshops and requirements analysis, including confirmation of approved data sources, access permissions, role-based access controls, audit requirements, infrastructure sizing, portal guidelines, and deployment constraints. EMI’s planned platform setup includes development, testing, pilot, and production environments, together with a local LLM runtime configured for the approved on-premises CPU infrastructure. |
The solution scope includes chatbot backend services, APIs, controlled tool and function calling, conversation and session management, audit logging, a portal-aligned user interface, and a vector database or approved local knowledge store. Data integration is planned for approved well and facility pressure data, production and deferment data, historical operational and surveillance data, and AVEVA PI System applications through approved interfaces. Retrieval-augmented generation and AI orchestration workflows are intended to ground responses in approved operational data, engineering references, formulas, troubleshooting guides, root-cause analysis reports, and operational best practices. The chatbot is also scoped to support trend analysis, comparisons with historical performance, anomaly and deviation identification, troubleshooting guidance, and operational decision support.
The project scope further includes on-demand visualizations and reports, export to approved business formats, testing and validation, pilot deployment, production deployment, training, documentation, and handover. Final outcomes and deployment status should be confirmed with the project team before publication.
