Start with reality
Physical systems have constraints, histories, tolerances, and failure modes. Useful AI begins there.
Applied Form explores intelligence for complex physical systems.
We are interested in the places where important decisions depend on partial data, expensive experiments, domain expertise, and the stubborn reality of matter.
01
AI systems that can reason with uncertainty, context, and constraint
02
Materials and manufacturing domains where data is scarce but consequence is high
03
Tools that help people make better decisions before perfect evidence exists
The terrain
Some of the most valuable domains do not produce internet-scale data.
Materials science, industrial R&D, sustainability, manufacturing, compliance, and deep technical services often work with sparse, uneven, proprietary, or expensive-to-generate evidence. The data is real, but rarely neat.
Applied Form is concerned with how AI can become useful in those conditions: not by pretending uncertainty disappears, but by helping experts hold more context, compare more possibilities, and decide what to learn next.
Our working stance
AI should respect the shape of the domain it enters.
Value sparse evidence
Small datasets can contain hard-won knowledge. The question is how to make that knowledge usable.
Keep humans in the loop
The best systems sharpen expert judgement instead of replacing the expertise that gives data meaning.
Build for consequence
When experiments are expensive and decisions matter, elegance has to include traceability and restraint.
Where we are looking
A company for the domains where intelligence has to be earned.
Materials and matter
Formulations, processing conditions, performance, sustainability, cost, and scale-up all interact. That makes materials a demanding and fascinating proving ground for AI.
Sparse data domains
Many important organisations cannot wait for perfect datasets. They need systems that work with partial evidence, expert input, and continuously changing context.
Decision support
Applied Form is drawn to tools that help teams ask better questions, see the assumptions behind choices, and move from ambiguity toward informed action.
Founder
Building Applied Form for domains where certainty is expensive.
I am Jordan. I work at the edge of technology, strategy, and applied science. For more than two decades, I have helped turn ambiguous problems into products, teams, and systems across creative, commercial, and technical environments.
Applied Form is where that experience meets the questions I care about now: how AI can support judgement when evidence is sparse, experiments are costly, and context matters.
The promise is not just speed. It is making hard-won knowledge easier to use without flattening uncertainty into false confidence.