Independent research · Birmingham, Alabama

Make possible what was not possible before.

Durablis Research investigates difficult technical questions whose answers can create new capabilities—not merely new products.

We do not begin with a feature, a roadmap, or a predetermined artifact. We begin with a question important enough that the answer would matter even if it never became a product.

Research programs

Three questions large enough to grow into fields.

These programs are deliberately broad. Each contains problems that can produce software, papers, protocols, experiments, failed hypotheses, or entirely new lines of inquiry.

Loading canonical research records…

Research notes

Write down what we learn.

Research notes are durable records of questions, hypotheses, experiments, evidence, conclusions, and changed minds. A useful negative result belongs here too.

Loading research notes…

Open problems

Problems we do not know how to solve yet.

These are invitations to think with us. They are intentionally more specific than the programs and less constrained than projects.

Loading open problems…

How we work

Question → hypothesis → experiment → evidence → conclusion → artifact.

01

Capability over output

Code, papers, datasets, and prototypes are artifacts. The goal is increased capability and understanding.

02

Evidence over theater

Activity is not progress. We want evidence that reality changed or that our understanding became less wrong.

03

Negative results count

Proving an attractive approach does not work can be more valuable than shipping another implementation of it.

04

Reality closes the loop

Research should encounter production systems, real constraints, real users, and real failure modes whenever possible.

About Durablis Research

A small lab with a long horizon.

Durablis Research is an experiment in building a modern research institution from the bottom up: start with important questions, preserve what is learned, connect research to real systems, and let the institution accrete around useful work.

We are intentionally early. The aim is not to imitate the scale of historical industrial laboratories. It is to recreate the conditions that made them productive: patient inquiry, unusually capable collaborators, cross-disciplinary work, and a tight feedback loop with reality.

Interested in one of these problems? Start in the research repository. The canonical program, note, and open-problem records are version-controlled there.