Glossary
Decision DNA
Decision DNA is a company's judgment captured as a structured, company-owned asset. It is built from a knowledge library of source files, the documents a company already has, plus enriched files that encode contexts and catalogs: the standards, precedents, and rules of thumb that senior experts carry in their heads. AI engines reason over it to produce decisions with source references.
Why does decision DNA matter?
Most companies run on the judgment of a few senior people. That judgment decides which standard applies when two conflict, what a past failure taught, and which client constraint overrides which default. Almost none of it is written down, and all of it leaves the building when the expert does.
Decision DNA turns that judgment into an asset the company owns. Documents alone are not enough: a folder of PDFs tells an AI engine what the company wrote, not how the company decides. The enriched files are the difference. They are created with the experts and encode the contexts and catalogs behind their calls, so the knowledge library reflects judgment, not just paperwork.
How is decision DNA used?
It is the knowledge half of a knowledge and control layer. When a question reaches the platform, the engines reason over the decision DNA rather than over the open internet or an undifferentiated document dump. Every output carries a reference back to the specific files it rests on, along with a calibrated confidence level, and the platform abstains when the decision DNA does not sufficiently support a conclusion.
Because decision DNA sits above the AI engines rather than inside one, it survives model churn. Engines are swapped as better ones appear; the asset, and everything the company invested in building it, stays put.
What decision DNA is not
It is not a chat history, a prompt library, or a fine-tuned model. Fine-tuning bakes knowledge into one vendor’s weights, where it cannot be inspected, corrected, or moved. Decision DNA stays outside the engines: inspectable, correctable, and owned by the company, which is what makes outputs built on it checkable.
See how Praxiron works for how decision DNA is built in practice, or read the category pillar, What is a knowledge and control layer.