Research & Open Source

Build the Future of AI for Energy

Join a Stanford PhD-led research network building open-source AI tools for the oil and gas industry. Contribute to real projects. Publish real research. Build a portfolio that matters.

Apply Below ↓

What We're Building

Groundwork Analytics builds AI tools and research for the upstream oil and gas industry. Our founder holds a PhD from Stanford in computational science and has spent over a decade at the intersection of petroleum engineering and machine learning.

We maintain petro-mcp (GitHub), an open-source MCP server that connects AI assistants to oilfield data. We publish technical articles at petropt.com that are read by engineers and operators across the industry.

We're looking for collaborators who want to work on real problems, build real tools, and contribute to something that matters.

What You Get

More Than an Internship

Mentorship from a Stanford PhD

Direct feedback on your code, writing, and research from someone who's published in top journals and built production AI systems.

Real Project Experience

Work on open-source tools used by real engineers. Your contributions ship to PyPI, not a shelf.

Published Work

Co-author technical articles on petropt.com and potentially conference papers. Your name on real publications.

Recommendation Letters

Strong contributors receive detailed recommendation letters for grad school or job applications from a Stanford PhD.

Industry Exposure

Learn how AI is actually deployed in oil and gas. Understand the domain, the data, and the business context.

Portfolio You Own

Everything you build is open source. Your GitHub contributions, articles, and projects belong to you forever.

Who We're Looking For

Ideal Collaborators

Petroleum Engineering Students

Undergrad or grad students who want to combine domain knowledge with AI and data science. You understand reservoirs, drilling, or production and want to build tools that apply to real workflows.

Computer Science & Data Science Students

You know Python, ML, or LLMs and want a domain to apply your skills. Energy is one of the most data-rich, underserved industries for AI. We'll teach you the domain.

Early-Career Engineers & Researchers

Recently graduated or in your first role. Looking for meaningful side projects, publications, or a transition into AI/data science. We offer structure and mentorship that most side projects don't.

Technical Writers

You can explain complex technical concepts clearly. We need well-written articles, documentation, and tutorials that help engineers understand and adopt AI tools.

Current Projects

What You Could Work On

petro-mcp

An open-source MCP server connecting AI assistants to oilfield data sources (state commissions, production databases, well logs). Built in Python, published on PyPI.

Needs: Python, API integration, data engineering, testing

Production Agents

AI agents that automate production engineering workflows: decline curve analysis, anomaly detection, allocation, and reporting. Built on LangGraph and Claude.

Needs: Python, LLM agents, production engineering knowledge

Decline Curve AI

Physics-informed machine learning models for production forecasting that combine Arps decline curves with neural networks. Research-grade work with publication potential.

Needs: ML/deep learning, physics-informed neural networks, reservoir engineering

Completion Analytics

Data analysis pipelines for completions and frac data in the Permian Basin. Statistical modeling, feature engineering, and visualization of treatment parameters vs. production outcomes.

Needs: Data analysis, statistics, Python, completions knowledge helpful

Technical Content

In-depth articles on AI applications in petroleum engineering. Research-backed, practitioner-focused content published on petropt.com and distributed across industry channels.

Needs: Technical writing, research skills, petroleum engineering or AI background

How It Works

Simple Process

1

Apply below

Fill out the short form. Takes about 5 minutes.

2

We review

We read every application and respond within one week.

3

Introductory call

A 20-minute video call to discuss your interests and find the right project fit.

4

Start contributing

Get onboarded to a project with clear tasks, documentation, and regular check-ins.

5

Work at your pace

Most collaborators contribute 5-15 hours per week. Fully remote, flexible schedule.

FAQ

Common Questions

Is this a paid position?
No. This is an unpaid research collaboration, similar to an academic research assistantship. You gain mentorship, published work, recommendation letters, and portfolio pieces. We're transparent about this upfront because we believe the experience and outputs are genuinely valuable.
How much time do I need to commit?
Most collaborators contribute 5-15 hours per week. We're flexible and understand you have other commitments. What matters is consistency and quality, not clocking hours.
Do I need oil and gas experience?
No. If you're a CS or data science student, we'll teach you the domain. If you're a petroleum engineering student, we'll help you build the technical skills. Curiosity and willingness to learn matter more than existing knowledge.
Is this remote?
Yes, fully remote. We collaborate via GitHub, Slack, and video calls. We welcome collaborators from anywhere in the world.
Can this count toward my thesis or capstone?
Potentially, yes. Several of our projects have publication potential and could align with thesis or capstone requirements. We're happy to work with your advisor to make it count.
What tools and technologies do you use?
Python, GitHub, Claude/LLMs, LangGraph, MCP protocol, pandas, scikit-learn, PyTorch, and various oilfield data APIs. You don't need to know all of these; we'll help you learn what's needed for your project.
How many collaborators do you take at a time?
We keep the group small to ensure quality mentorship. Typically 3-5 active collaborators at any given time. We'd rather work closely with a few people than spread thin across many.
What happens after the collaboration ends?
You keep everything: your code contributions, articles, and portfolio. Strong collaborators get recommendation letters and stay part of our network.

Join Us

Apply Now

Takes about 5 minutes. We review applications on a rolling basis and respond within one week.

Typical collaboration: 3-6 months, 5-10 hrs/week, fully remote.

Questions before applying?

Email info@petropt.com or book a 15-minute call.