KI-Start-Ups in der Pharmaindustrie: Gründungscluster, Schwerpunkte und Nischen
Bibliographic record
Abstract
Im vorliegenden Beitrag werden KI-Start-Ups aus der Pharma- und Gesundheitsbranche analysiert. Um weltweite Gründungscluster sowie die Arbeitsschwerpunkte der Unternehmen in diesen Clustern aufzudecken, werden Unternehmensbeschreibungen aus der Crunchbase-Datenbank mithilfe einer Clusteranalyse ausgewertet. Die Ergebnisse der Clusteranalyse zeigen, dass die technologische Entwicklung von Start-Ups aus dem anglo-amerikanischen Raum dominiert wird. Unternehmen aus den USA, Großbritannien und Kanada arbeiten dabei schwerpunktmäßig an der Entwicklung von KI-basierten Verfahren für die Arzneimittelforschung. Start-Ups aus Großbritannien engagieren sich zudem in der Entwicklung intelligenter Data-Management-Systeme für den Gesundheitssektor. Vor dem Hintergrund der starken internationalen Konkurrenz wird für deutsche Unternehmen die Verfolgung einer Nischenstrategie empfohlen. Mit Blick auf die KI-gestützte Entwicklung von Biomarkern wird dabei ein Anwendungsfeld mit echtem Aufhol- und Anschlusspotenzial identifiziert, in dem die internationale Konkurrenz noch nicht stark vertreten ist. In this paper, AI startups from the pharmaceutical and healthcare industries are analyzed. In order to uncover global startup clusters and the focus of work in these clusters, company descriptions from the Crunchbase database are evaluated using a cluster analysis. The results of the cluster analysis show that technological development is dominated by start-ups from the Anglo-American region. Companies from the USA, the UK and Canada are focusing on the development of AI-based technologies for drug discovery. Start-ups from the UK are also involved in the development of intelligent data management systems for the healthcare sector. Against the background of strong international competition, it is recommended for German companies to pursue a niche strategy. With regard to AI-based development of biomarkers, this study identifies a field of application with real potential for catching up and connecting, in which the international competition is not yet strongly represented.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.018 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.009 |
| Science and technology studies | 0.004 | 0.001 |
| Scholarly communication | 0.008 | 0.004 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.022 | 0.006 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".