Lantern Pharma (NASDAQ:LTRN) showcased a live demonstration of its withZeta.ai platform during a RedChip Companies webcast, positioning the system as both an internal drug-development engine and a potential subscription business for external users. The session was introduced by Paul Kuntz, Communications Director at RedChip Companies, who said the event would provide a “live real-time demonstration” of the company’s “next generation AI platform designed to transform how oncology drugs are discovered, particularly in rare cancers.”
Lantern Pharma CEO, President, and Director Panna Sharma said withZeta began as an internal tool to help the company determine which cancer indications to pursue and which drug combinations might be most promising. Sharma traced the platform’s origins to a question he received after presenting on Lantern’s use of data in rare disease research, noting the company “ended up with six orphan drug designations and four rare pediatric disease designations and two fast track,” which prompted him to consider whether Lantern could extend its approach to other drug developers.
Rare-cancer focus and curated dataset
Sharma said the platform is designed to perform tasks ranging from real-time literature review and pathway analysis to predicting drug properties such as blood-brain barrier penetrability and ADMET-related characteristics. He also described an ambition to create a “living knowledge graph” that packages and shares findings across teams, arguing that the system can compress work that might take “hours, days, weeks” into “minutes.”
Modes, personas, and transparency
During the demonstration, Sharma outlined multiple usage modes intended for different levels of depth and speed. He described an “investigator mode” for deep work, an “explorer mode” for faster conversational querying, and a “reporter mode” for compressing extensive research into a shareable report.
He also said withZeta includes selectable “personas” aligned with roles in drug development—such as medicinal chemist, clinical trial strategist, clinical oncologist, and biomarker specialist—and that these personas can interact rather than operating as isolated agents. In future versions, Sharma said the company expects to support “teams of these co-scientists.”
Sharma emphasized transparency in how the platform reaches conclusions, saying it shows which tools it uses and how information is integrated so the process “is not a black box.”
Live workflow: literature synthesis, ranking, and knowledge graphs
In a live query, Sharma asked: “What rare blood cancers are in high patient need in children?” He said the system rapidly digested dozens of sources across PubMed publications and clinical trials. He highlighted interface features such as real-time entity highlighting—genes in green and diseases in red—intended to help drug developers scan outputs quickly.
Sharma then switched to a medicinal chemist persona and asked the system to discuss challenges with menin inhibitors and to rank preclinical candidates in Phase II and III trials. He said the output addressed issues such as differentiation syndrome and resistance, and it produced a tiered ranking, including a “Tier 1” entry for an approved drug and a “phase II registrational trial” entry identified as KO-539, which the system described as having a regulatory submission anticipated, while also noting limitations including “very limited pediatric data.”
As the session progressed, Sharma pointed to withZeta’s knowledge graph, showing relationships among drugs, diseases, and genes. He said the graph can be exported as an interactive HTML file or as machine-readable JSON for enterprise use and knowledge retention.
Generative chemistry example and budgeting output
Sharma asked withZeta to help design a novel menin inhibitor with better combination potential and fewer side effects. He described the system using multiple tools and sources and iterating through molecule-generation attempts, including rejecting early designs that did not pass drug-likeness filters. Sharma said the platform produced a final design with a molecular structure and SMILES string and compared it against prior iterations, providing predicted characteristics and a proposed testing and development path.
He also prompted the system to generate a budget and aggressive schedule to develop the molecule from “research grade to full GMP,” arguing that such work often takes weeks when gathering estimates from contract research organizations. Sharma said Lantern has trained withZeta using the company’s experience taking drugs “from ideas on a whiteboard” through manufacturing and into clinical dosing, and he said that internal experience is embedded in the tool’s outputs.
Q&A: failed trials analysis and collaboration roadmap
In response to a question about whether withZeta can analyze failed or terminated trials, Sharma initiated a prompt asking it to review “recent phase II failures in rare cancer clinical trials” and explain limitations. He said the system returned a structured analysis that included trial design flaws, enrollment barriers, and manufacturing bottlenecks in cell therapies, along with recommendations such as basket trials. Sharma highlighted a conclusion from the output that such failures “are not random events, but reflect systematic mismatches between trial assumptions and rare disease realities.”
Sharma also said Lantern is using withZeta internally and reported “probably now 50 going on 100 external users,” adding that adoption is “growing pretty rapidly.” Addressing collaboration, he said an enterprise feature set is planned, including team workspaces, collaborative annotation, multi-user investigation sessions, credentialed researcher profiles, personalized feeds, white-labeling, and API access.
Looking ahead, Sharma described a broader vision for withZeta to support additional data modalities, deeper pathway knowledge, improved biology models, mobile optimization, and tools for IND preparation and filings. He characterized the long-term goal as creating “almost like the Bloomberg for medicine,” starting with oncology and potentially expanding beyond cancer.
About Lantern Pharma (NASDAQ:LTRN)
Lantern Pharma, Inc is a clinical-stage oncology company leveraging artificial intelligence (AI) and machine learning to accelerate the discovery and development of targeted cancer therapies. Headquartered in Dallas, Texas, Lantern Pharma’s proprietary RADR® platform integrates large-scale genomic, transcriptomic and chemical data to identify novel drug candidates and predict patient populations most likely to benefit from treatment.
The company’s pipeline focuses on molecules designed to address cancers with high unmet medical need.
