Every publication on this platform takes the same route: from the question through the source search and the AI-assisted analysis to professional review. We lay this route open — including what the AI does not take on.
At every step it says what the artificial intelligence takes on — and what it expressly does not take on.
It starts with a precise question. It comes from practice: from what therapists describe to us, and from what has remained unresolved in the literature.
Breaking the question down into searchable sub-aspects, including technical terms, synonyms and English equivalents.
The choice of question. What gets researched is decided by people — not by a model and not by search volume.
The search is systematic, across international study platforms and qualified specialist sources. The inclusion and exclusion criteria are fixed beforehand and documented alongside.
Searching large volumes of literature and capturing title, abstract, study type, sample size and year of publication.
The decision which source counts. Criteria are set in advance so that the selection stays verifiable and is not adjusted afterwards to fit.
Across many papers, recurring endpoints, divergent findings and methodological differences are made visible. This is exactly where the benefit lies: in a volume no individual can survey any more.
Summarising, structuring and comparing. Also finding papers that contradict each other.
The verdict on efficacy. It reproduces what was reported — not whether it is true. We carry this distinction through the entire presentation.
In this step every statement is checked against the original source: does the summary match the paper? Are limitations named? Is counter-evidence missing? Whatever does not hold up is corrected or discarded.
Nothing. It lays the references side by side so the review can be done quickly — no more than that.
The approval. Without professional review, no summary is published here.
Published with the date, the full list of sources and a note that the summary was produced with AI assistance. Every statement can be traced back to the primary source.
Formatting, keywording and linking to related topics in the library.
The labelling itself. That something was produced with AI assistance is always stated — that is a rule, not a setting.
On humic acid complexes and gut health there is one randomised, double-blind, placebo-controlled study in diarrhoea-predominant irritable bowel syndrome, together with investigations on human tissue and a broad preclinical basis. Three points of action are described above all: the tightness of the gut barrier, the binding of substances in the gut lumen, and inflammation-associated markers. A clinical review in Frontiers in Pharmacology places the findings in context for use in humans. It should be noted that the origin and processing of the materials studied differ considerably — a result applies first of all to the material with which it was obtained. The number of controlled human trials is still small; they point in the same direction and are growing.
This summary shows the form: finding, type of investigation, source — and at the end, what remains open. Every answer from the research looks like this.
Established, plausible or hypothesis — every statement carries its grade. With us a hypothesis is not a flaw but an invitation to test it.
We show the chain — from stress via cortisol to inflammation, from the gut barrier to the microbiome — and name the point where research is still missing.
We do not rate commercial preparations and make no purchase recommendation.
No patient data is processed or stored on this platform.
New studies and research focuses, as soon as they appear.