AppliedXL

AppliedXL is now leveraging its fact-checking and verification technology to resolve prediction market contracts on Kalshi.

AppliedXL is partnering with Kalshi to provide independent resolution analysis for biotech and pharmaceutical contracts The markets are designed to make expectations around clinical trial outcomes and FDA decisions more visible, while recognizing their potential impact on patients, researchers, physicians, and investors. AppliedXL evaluates outcomes against predefined public sources, including trial registry data, FDA documents, and company disclosures. Kalshi remains the final adjudicator under its exchange rules. The credibility of these markets depends on clear rules, reliable evidence, and strong safeguards against conflicts of interest, manipulation, and misuse. That is the verification layer we are helping build. Alongside the launch, we are publishing a joint report with Kalshi, Biopharma’s Public Probability, examining the potential benefits, limitations, and ethical risks of prediction markets in drug development, including insider trading, patient enrollment, and market design. https://lnkd.in/gM2aPX8n

For two hundred years, information companies did two jobs: find out what's true, and deliver it. AI just made the second one free. That single shift moves all the value to the first job: being first to know something. We call it Original Intelligence, and we think it's how news and information companies win from here. Our new report lays out the thesis and what it means for your business: → Delivery is now worth close to zero. Retrieving, summarizing, formatting. Machines do it at no cost. → Origination is worth more than ever. Producing a verified fact that didn't exist until you produced it, from primary records no one else has connected. → The scarce thing isn't the pipes. It's the water. The report is interactive, and it ends with a short assessment that maps where your business already holds original ground and where AI is commoditizing you. Read it, download the PDF and dataset, and run the assessment for your company here: https://lnkd.in/gNnxdm9p

AI isn't just writing the news. It's reading it first. AppliedXL's Francesco Marconi sat down with EIN Presswire to talk about the pre-news window and what that means: for journalism, for communications, and for anyone whose job depends on information reaching the right audience. https://lnkd.in/ej7tdf9k

Biotech is the highest-volatility corner of public markets. A single trial readout or FDA decision can move a stock 50% in a session. For decades, quantitative investment strategies have attacked that volatility with one signal: the science. Will the drug work? That's half of the question. A great drug in a poorly run trial still fails. Enrollment collapses. Timelines drift. Protocols get rewritten mid-study. Operational execution predicts outcomes as reliably as the molecule, yet it leaves almost no trace in the models that move capital. The data exists. It is public. It is federally mandated. It is updated in real time across 500,000+ clinical trials. It has simply never been read systematically. AppliedXL was built to read that signal. When we used it to simulate a quantitative trading model, the results were revealing: while the biotech sector lost 13% over the same period, the model returned +204%. The alpha is hiding in plain sight, if you know where to look. Full research below. 👇 https://lnkd.in/eg4xyQ3a

AppliedXL's submission to the FDA Drug Repurposing Docket (FDA-2026-N-4492). In May 2026, the FDA opened a public docket soliciting input on approved drugs with clinical evidence for new indications where no commercial sponsor has reason to pursue approval. It is a structural problem: once a drug loses patent protection, the regulatory pathway to a new labeled use has no obvious funder. Using our clinical trial intelligence platform, we identified five candidates with documented evidence and no active sponsor: → Gabapentin for menopausal vasomotor symptoms → Topiramate for alcohol use disorder → Pioglitazone for dementia prevention in post-stroke patients → Clomiphene for male hypogonadism secondary to obesity → Metformin for cancer prevention in T2DM The drugs exist. The evidence exists. The regulatory pathway is defined. What is absent is the mechanism to move one toward the other. https://lnkd.in/ej6pG5-8

We just published new research on clinical trial prediction, and the results set a new state of the art. 87.3% directional accuracy on endpoint prediction. A 17-point gap over the best published model in the literature. Under walk-forward temporal validation: 0.91 on regulatory approval, beating the Novartis DSAI benchmark. The key insight: most models ask whether a trial succeeds or fails. Ours decomposes the trial lifecycle into five conditional stages, each with a distinct failure mechanism and feature set. Combined with a domain-specific LLM extraction pipeline and a temporal LSTM that updates at every public announcement, the architecture produces a fundamentally different class of signal. Full paper out now. Link in comments.

I'm super proud of this one. We built our Assistant to provide institutional-grade clinical intelligence that beat anything else available, and now we have the numbers to prove it.

Today we're releasing our first public benchmark on vertical AI performance in biopharma intelligence. Five systems, ten clinical intelligence queries, four independent raters, self-preference bias corrected. AppliedXL outperformed Claude Opus, ChatGPT 5.4, and Perplexity Pro on depth, domain specificity, signal quality, and actionability — ranking first on all ten queries. Our floor score (46.5/50) exceeded every general-purpose system's ceiling (35.0/50). Full methodology and per-query results published for replicability.