Source: Unite.AI
XtalPi Holdings Limited announced the debut of Kodexia on October 6, 2026, describing it as the industry’s first closed-loop small interfering RNA (siRNA) discovery platform to combine generative AI, first-principles-driven biological modeling, and automated experimentation, and simultaneously disclosed six proprietary therapeutic programs built on the platform.
The company presented Kodexia as the expansion of its technology into nucleic acid therapeutics. The release describes siRNA therapeutics as a modality that silences disease-driving genes upstream of protein production, offering potential for biological targets that conventional modalities struggle to address. It lists the development constraints the platform is intended to answer: fragmented sequence and modification design processes, restrictive patent barriers, inconsistent translation from cellular assays to animal models, and limitations in delivering such therapies beyond the liver.
The siFormer Architecture
To address those constraints, XtalPi said it developed siFormer, the proprietary architecture at the core of Kodexia. According to the company, siFormer integrates RNA interference mechanisms with nucleic acid chemistry principles and incorporates RNA thermodynamics and structural features relevant to strand loading, target accessibility, and silencing efficiency. XtalPi said these mechanistic constraints steer molecular generation toward biologically plausible designs rather than relying on sequence correlations alone.
Through this unified approach, the company said, Kodexia is built to let researchers systematically balance potency, in vivo translation, durability, off-target activity, safety, and patentability rather than optimizing single endpoints in isolation. XtalPi’s Kodexia product page states that the platform applies AI across target-region discovery, sequence design, modification recommendation, experimental validation, delivery optimization, and intellectual property layout, pairing integrated sequence-and-modification design with a model built to predict in vivo efficacy early in development.
Closed-Loop Operations and Reported Performance
To execute that multiparameter optimization, XtalPi said, Kodexia runs a closed-loop process of computational design, automated experimentation, and empirical feedback. Generative models and the siFormer architecture operate continuously with the company’s automated laboratories, which conduct more than 500 in vitro experiments and 30 in vivo experiments each week.
XtalPi reported that this integration supports the parallel development of multiple target programs and has produced a nearly threefold increase in molecular design efficiency compared with conventional workflows. The company also reported that more than 50% of first-round designs tested across multiple programs showed stronger in vivo activity than positive controls. The product page states a design-efficiency gain of two to three times over traditional research and development, and describes experimental results feeding back into model iteration as programs advance.
Dual-Target Design and Delivery Beyond the Liver
XtalPi said Kodexia globally co-optimizes dual-target siRNAs as single molecules, a strategy the company describes as suited to disease biology that may not be adequately addressed through single-target intervention. For delivery beyond the liver, the platform co-designs siRNA molecules with their targeting systems and applies multi-objective optimization to antibody, peptide, and small-molecule conjugates.
The company said these delivery capabilities are being applied to research targeting the kidney, spleen, and adipose tissue. The product page additionally lists lipid nanoparticles among the platform’s delivery approaches and notes that XtalPi is developing AI prediction models for antisense oligonucleotides and other oligonucleotide modalities alongside its siRNA work.
Pipeline Status and the Lead Program
Kodexia currently drives six proprietary programs across metabolic, renal, respiratory, and central nervous system diseases, with more than half having completed in vivo efficacy evaluations, according to the company.
The lead program, in immunoglobulin A (IgA) nephropathy, reached non-human primate efficacy data within seven months. XtalPi said the candidate demonstrated greater activity and durability than a clinical-stage reference molecule directed at the same target in a parallel preclinical comparison. The program is on track for preclinical candidate selection within a nine-month project timeline; the release contrasts that pace with what it describes as an industry norm of 12 to 18 months. XtalPi said each internal program generates experimental data that feeds directly back into the platform’s models.
Company Background and Prior Disclosure
XtalPi Holdings Limited was founded in 2015 by three physicists from the Massachusetts Institute of Technology. The company provides AI- and robotics-driven research and development solutions, built on quantum physics, artificial intelligence, high-performance cloud computing, and standardized automation systems, for the pharmaceutical, materials science, agricultural technology, energy, new chemicals, and cosmetics industries. It describes Kodexia as extending a drug-discovery platform that already spans small molecules, antibodies, and peptides.
The October 6 debut was not Kodexia’s first public appearance. XtalPi presented the platform’s pipeline progress at the CPIC 2026 conference, according to a company post dated July 26, 2026, which said the IgA nephropathy program had reached non-human primate efficacy data in seven months and was expected to confirm a preclinical candidate molecule within nine months of project initiation. That post described five differentiated siRNA programs in development at the time, across metabolic, renal, dual-target, and extrahepatic-delivery directions; the October 6, 2026 release states that the platform currently drives six. XtalPi said it is advancing those internal assets alongside industry collaborations and is pursuing platform partnerships, asset co-development, and out-licensing arrangements with pharmaceutical and biotechnology companies.