Target Safety Assessment in 1 day
Ensuring the path to a safe cure for
cancer
is never lost in discovery
AutoTSA reads thousands of scientific papers, extracts the safety evidence that matters, and delivers a complete target safety assessment — fully automated, verifiably trustworthy.
Automation
From 1000+ papers to safety assessment. One click.
Point AutoTSA at a target and let it work. The entire pipeline — from literature ingestion to risk-ranked report — runs autonomously so you can start with the first version of the end report.
Our method
Start with the full end report.
AutoTSA uses a multi-agent pipeline built specifically for scientific evidence extraction. Rather than prompting a single model to summarize, separate agents read, extract, verify, and synthesize — and we validate results against human expert benchmarks.
Structured interpretation
Verbatim quotes are extracted from full-text papers and augmented with evidence type, severity, and safety category.
Verifiable
Every claim links back to its source text. We verify that claims are actually supported by the evidence.
Benchmark-validated
We evaluate against expert-curated safety claims and measure recall, precision, and verification quality for every pipeline change.
Evidence extraction
Every claim traces back to its source
AutoTSA extracts safety-relevant quotes from scientific literature and maps each finding to its evidence base — so you can always verify what the AI found.
"IL-11 signaling through IL-11Rα promotes cardiac fibrosis and contributes to adverse ventricular remodeling following myocardial infarction in both murine models and human tissue samples."
Schafer et al., 2017 — Nature
Extracted quotes
Structured output
Safety claims organized by category
Extracted evidence is synthesized into actionable safety claims, organized by organ system and risk category — the same structure toxicologists expect in a TSA.
IL-11 promotes cardiac fibrosis through activation of cardiac fibroblasts and TGF-β signaling
Anti-IL-11 treatment reverses established cardiac fibrosis in murine pressure-overload models
Elevated IL-11 correlates with adverse cardiac remodeling post-myocardial infarction
Genetic ablation of IL-11 signaling protects against diet-induced steatohepatitis
IL-11 drives hepatocyte death through ERK-dependent non-canonical pathway
IL-11 modulates macrophage polarization toward a pro-inflammatory phenotype in tissue injury contexts
Therapeutic neutralization does not impair immune defense in infection models
IL-11 is critical for decidualization and implantation; inhibition may impair fertility
For whom
Built for teams who can't afford to miss a safety signal
AutoTSA is our response to how LLMs change discovery. Our current focus is maximally improving automated TSA quality and are looking for partners to collaborate on this goal.
Pharmaceutical companies
Large pipelines of targets requiring continuous safety assessment as new literature emerges.
Biotech startups
Need rigorous safety evidence fast — without building a large internal toxicology team.
Contract Research Organizations
Deliver faster, more consistent assessments while maintaining full regulatory traceability.
About
Built by experts in toxicology and AI
Longtail Discovery combines over 20 years of pharmaceutical target safety assessment experience. Our team comes from TNO — the Netherlands Organisation for Applied Scientific Research — and brings together domain expertise and generative AI.
Dr. Jennifer McCormack-Venhorst
Lead
>20 years experience in pharma. TargetTri lead — proven TSA platform.
Noah Smeets
AI Expert
>5 years strategy and innovation consulting. AI solution design expertise.
Dr. Tanja Rouhani Rankouhi
Toxicologist
>20 years toxicology and TSA expertise. TSA software user. Deep TSA process understanding.
Hein Kolk
AI Expert
Expertise in Semantics in Generative AI. Experience at start-ups as developer.
Get in touch
Request a demo
See AutoTSA in action on a target relevant to your pipeline.