The rare disease community has nonprofits. It has research databases. It has government registries. What it does not have is an intelligence layer, and the absence of that layer is the reason so many people across this community still manually check five or six websites every week to understand what changed for their disease.

This is not a criticism of what exists. NORD, ClinicalTrials.gov, PubMed, NIH RePORTER, and FDA's orphan drug databases are serious institutions doing important work. The problem is structural. Each was built for a specific purpose, and no amount of improvement to any one of them will close the gap between them. That gap is a missing category, not a missing feature.

Three categories that exist today

If you map the current rare disease information landscape, the major players fall into three categories. Each one is well-defined and does its job.

Large nonprofits focused on advocacy and education. NORD, Global Genes, the EveryLife Foundation, and hundreds of disease-specific PAGs. They produce curated educational content, advocate for policy, connect patients with resources, and build community. Their value is in curation, trust, and direct support. They are not, and should not have to be, real-time data monitoring systems.

Government databases built for compliance and archival. ClinicalTrials.gov, PubMed, NIH RePORTER, FDA's various drug and designation databases. These are authoritative record systems. ClinicalTrials.gov exists because federal law requires trial registration. PubMed exists to index the biomedical literature. NIH RePORTER tracks how taxpayer-funded research dollars are allocated. They are designed for completeness and accuracy, not for answering the question "what changed this week for Duchenne muscular dystrophy across all domains?"

Research and clinical tools built for specialists. Orphanet, OMIM, GeneReviews, ClinVar. Deep, detailed, and essential for clinicians and researchers who need genetic, phenotypic, and clinical reference data. A genetic counselor uses these daily. A biotech team building a competitive landscape does not.

Each category is valuable. Each is well-served by existing institutions. And none of them, individually or in combination, provides what a growing number of people in rare disease actually need: a continuously updated, cross-referenced, structured view of everything happening around a specific disease.

The clinician tracking emerging therapies needs it. The researcher identifying evidence gaps before writing a grant needs it. The PAG director reporting to a board needs it. The biotech team evaluating a therapeutic area needs it. The foundation officer deciding where to allocate funding needs it. The parent trying to understand the landscape for a newly diagnosed child needs it. They all need the same thing. None of them have a tool built for it.

The missing category

The missing category is intelligence. Not artificial intelligence (though AI plays a role in building it). Intelligence in the way an analyst at a pharmaceutical company uses the word. Structured information from authoritative sources, monitored over time, with changes detected and contextualized across domains.

Here is what that means in practice. A clinical trial for a rare disease moves from Phase 2 to Phase 3. That fact exists in ClinicalTrials.gov. But to understand what it means, you also need to know: Does this drug have an orphan designation? (FDA database.) What does the published evidence look like for this intervention? (PubMed.) Is NIH funding active research on the same mechanism? (NIH RePORTER.) Has any relevant legislation been introduced? (Congress.gov or state legislature sites.) Are other companies working on competing therapies? (Back to ClinicalTrials.gov, but a different search.)

Today, the person who connects those dots is a human being with a browser and a spreadsheet. Sometimes it is a medical affairs team with resources. More often it is one person at a small organization doing it between ten other responsibilities. That model works for one disease if you are dedicated and methodical. It does not scale to the more than 10,000 rare diseases that need this attention.

The intelligence category is defined by four properties that no existing category provides:

Continuous monitoring with change detection. Not a static snapshot. A system that watches sources and tells you when something changed.

Cross-domain linkage. Trials connected to designations connected to publications connected to funding connected to policy. Not six separate searches.

Structured, classified data. Not flat lists of search results. Publications classified by research type. Trials grouped by status and phase. Grants organized by mechanism and theme. The structure itself is information.

Disease-anchored identity. One disease, one identifier, consistent across every data source. This sounds basic. It is the hardest infrastructure problem in the stack and the reason cross-referencing is so painful to do manually.

Why now

Two things changed that make this category possible now in a way it was not five years ago.

First, the data sources matured. ClinicalTrials.gov's API is comprehensive. PubMed's E-utilities are powerful. NIH RePORTER provides structured grant data. FDA publishes orphan designation and approval data in accessible formats. The raw material is available. What was missing was the integration.

Second, AI classification reached the point where it can reliably structure unstructured data at scale. Classifying 500 PubMed articles into research types, determining whether an NIH grant is a primary match or adjacent to a disease, flagging when a trial record changed in a meaningful way. These are tasks that required a human analyst five years ago and can now be done programmatically with verification. The key word is "with verification." The AI is a tool in the pipeline, not a replacement for authoritative data. Every claim traces back to a primary source.

Where KISHO fits

I built KISHO to be this missing category. It is a rare disease intelligence platform that indexes trials, publications, grants, regulatory actions, policy developments, and news across more than 10,000 rare diseases, anchored to the MONDO ontology so that cross-referencing actually works.

The free tools (clinical trials, publications, and grants for any disease, no login required) provide the structured snapshot that government databases don't. The platform (workspace, AI reports, saved searches, change alerts) provides the continuous monitoring that no existing tool does.

KISHO does not produce original research. It does not do advocacy. It does not curate patient education content. It reads from the same authoritative sources that already exist, structures what it finds, links it across domains, and watches it over time. The nonprofits, government databases, and research tools are the foundation. KISHO is the connective tissue.

What this means

If you work in rare disease, in any capacity, you have probably built your own version of the five-tab workflow. You check your sources, you track what changed, you try to hold the full picture in your head or in a spreadsheet that only you understand.

That is the work KISHO was built to do. Not instead of the institutions you rely on. On top of them. The rare disease community does not need another nonprofit, another database, or another research portal. It needs the layer that connects them. That layer is what intelligence means in this context, and it is the category that did not exist until now.