Founding Research Advisor — Remote, AI Group Decisions in Oxford

Overview

Help explore one of AI's most overlooked questions. Can AI help groups reach better decisions together?

This role exists because of one central question that has fundamentally changed how we think about SplitStay. The context for everything below is described in a founder post on LinkedIn.

If you have two minutes, I encourage you to read it first. It provides the context for everything below.

Why this work matters

Over the past year, building SplitStay has changed how I think about AI. When I started the company, I thought we were solving a problem in travel. Today, I think travel is simply where a much broader challenge becomes impossible to ignore.

Modern AI has become remarkably capable of helping individuals — search, write, analyse, translate and generate with extraordinary speed — yet helping a group of people navigate different preferences, constraints and trade-offs remains surprisingly difficult. That feels less like a software problem and more like a human one.

The hypothesis

I have a hypothesis: recent advances in AI create an opportunity to revisit decades of research into human coordination. Not to replace human judgement. Not to automate consensus. But to build tools that help groups understand one another better, explain trade-offs more clearly, reduce unnecessary friction, and ultimately help people reach better decisions together. I don’t know whether that hypothesis is correct, but I think it’s worth testing.

Where research meets software

Why I believe research belongs at the centre: questions like these have been explored for decades — Behavioral Science, Negotiation, Human–Computer Interaction, Collective Decision-Making, Organisational Behaviour, Computational Social Science, Multi-Agent Systems. The challenge isn’t that we lack research; it’s that very little of it reaches the software people use every day. I’d like SplitStay to become one place where those worlds genuinely meet.

Why travel

A group trip forces people to reconcile: before anything can happen. Booking technology already exists. Helping the group reach the booking decision is a different challenge. That is where we are focused.

The collaboration I’m looking for

I'm intentionally looking for one long-term collaborator — not someone to validate my thinking, but someone to challenge it. Someone who enjoys asking difficult questions and is excited by seeing rigorous research tested with real users solving real problems. Whether you’re a professor, postdoctoral researcher, research fellow or practitioner matters far less than how you think.

What success looks like

Success isn’t another startup feature or another academic publication. Success is building products that quietly embed decades of behavioural research into everyday life. If users coordinate more effectively without ever knowing why, that’s a meaningful form of research impact.

• Long-term strategic collaboration
• Approximately 20 hours per month
• Equity through the Founder Institute FAST Agreement
• Remote
• Direct collaboration with the founder

The structure matters far less than finding the right intellectual fit.

A personal invitation

If your work intersects with this question, I’d be delighted to exchange ideas. Not because I expect immediate agreement — quite the opposite. I’m looking for someone who enjoys questioning assumptions, because I suspect the most valuable insights will come from discovering where our current thinking is incomplete. If, while reading this, one particular researcher immediately came to mind, I’d genuinely appreciate an introduction.

Some collaborations begin with funding. Some begin with publications. Sometimes they begin with a single conversation. Perhaps this could be one of them.