Dr. Georg Heiler
Co-Founder / CTO
About Jubust
From patent noise to reviewable SAR evidence.
Jubust turns dense pharma patent PDFs into structured compounds, normalized bioactivity, assay context, and source-linked evidence that scientific reviewers can check.
Our thesis
Patent search is not the bottleneck. Verified SAR reconstruction is.
The hard work is connecting compound structures, assay values, units, cell lines, examples, and claims into data objects that scientists can verify and export into real workflows.
Now
Verified SAR datasets from pharma patent PDFs.
Next
Richer assay normalization and source-review workflows.
Later
Broader patent-intelligence coverage once SAR quality is strong.
Our team
We combine product, AI, infrastructure, and life-science software experience, starting narrow so the output is useful before the product surface broadens.
Manuel Bichler
Co-Founder / CEO
Advisors
Independent experts who sharpen our scientific method, benchmarking, and strategy as we build.
Univ.-Prof. Dr. Allan Hanbury
Technical Advisor · TU Wien, Data Intelligence
Allan Hanbury is Professor for Data Intelligence at TU Wien, working across information retrieval, text classification, and information extraction. As co-founder of contextflow and coordinator of major medical search and information-extraction projects, he brings practical judgment on retrieval quality, evaluation, and benchmarking to Jubust's source-grounded patent intelligence.
View profile
Prof. Dr. Giulio Superti-Furga
Scientific Advisor · Cori Institute, MedUni Vienna
Giulio Superti-Furga is a molecular and systems biologist whose work spans chemical biology, drug action, SLC transporters, and translational drug discovery. As a co-founder of Cellzome, Haplogen, Allcyte, Proxygen, and Solgate, he brings rare founder-scientist judgment to Jubust's work: extracting verified SAR from patent PDFs with the chemistry, assay context, and source evidence intact.
View profileWhat we optimize for
A dataset is only useful if a scientific reviewer can check it, trust it, and carry it into the next decision.
Reviewability
Important fields stay anchored to source evidence so reviewers can confirm what the system found.
Scientific rigor
Chemistry-aware validation, unit checks, and document consistency matter more than demo-friendly output.
Deterministic workflows
Replayable, versioned processing and review steps make quality changes deliberate.
Human-in-the-loop accountability
Uncertain fields should surface for review and correction instead of passing through with false confidence.
Shape the workflow with us
Request a sample dataset from a patent slice your team cares about, and we will show what verified SAR intelligence can look like.
- Source-linked outputs
- Expert review where it matters
- Dataset-ready exports