Agentic AI Workflows for Production and Reservoir Engineering

Published at Aug 19, 2025

Value Added: Enhanced analysis speed and effectiveness, streamlined workflows, foundation for future model improvements
Skills Used: Python, MCP, SvelteKit, OpenAI Agent SDK, Data Engineering
#AI Agents#Generative AI#Reservoir Engineering#Production Engineering#MCP Servers#SvelteKit

Summary

Currently exploring the development of MCP servers and agentic AI workflows to enhance the effectiveness and speed of production and reservoir engineering analysis. The goal is to streamline well analysis and empower engineers with intelligent, context-aware tools. A SvelteKit-based front end is being created to improve user experience and facilitate context gathering for AI agents.

Project Highlights

  • Building MCP servers with tools for production and reservoir engineering analysis
  • Designing agentic AI workflows to automate and accelerate well analysis
  • Developing a SvelteKit front end for intuitive user interaction and context collection
  • Leveraging personal resources and public data from the AOGCC for initial development
  • Utilizing OpenAI’s agent SDK to enable advanced AI capabilities

Technical Innovation

  • Creation of a robust evaluation set to measure model performance and reliability
  • Planning for future fine-tuning to enhance agent abilities for petroleum engineering analysis
  • Considering model distillation to maintain domain applicability while reducing invocation costs
  • Integrating public data sources and personal infrastructure for rapid prototyping

Impact

This ongoing work lays the foundation for agentic AI workflows in petroleum engineering, with the potential to significantly improve analysis speed, accuracy, and accessibility. The combination of MCP servers, SvelteKit front end, and advanced AI agents will enable engineers to streamline their work and unlock new capabilities for well and field analysis.

Rodney Murray © 2026