AI has made exploration abundant.
Verified scientific discovery remains scarce.
An AI-native scientific discovery company turning computational exploration into patentable material IP assets.
Lymeric broadly explores hypothesis, literature, composition, and process spaces with autonomous AI agents, verifies high-potential candidates through multi-fidelity physics simulations and selective laboratory experiments, and secures the verified results as proprietary material IP assets.
Rather than selling software seats, we deliver experimentally proven material assets—starting with next-generation battery materials.
Agents explore
Hypotheses, literature, and composition spaces
Physics filters
Multi-fidelity simulations and surrogate screening
Experiments prove
Selective pilot synthesis and lab validation
Lymeric owns
Composition, process, and application material IP
Thesis
Science is shifting from attention-limited to verification-limited.
AI agents are beginning to make meaningful progress in domains where verification is digital, fast, and inexpensive—such as code synthesis, formal mathematics, and theoretical modeling. In these digital domains, candidate solutions can be validated in milliseconds through logical execution and computation.
In physical and materials sciences, however, digital hypothesis generation is only the starting point. Physical verification remains scarce, slow, and expensive. Converting a candidate into a defensible material asset requires chemical synthesis, sample and electrode fabrication, long-duration cycle-life testing, degradation root-cause analysis, batch reproducibility confirmation, and customer-specific qualification.
The bottleneck in modern scientific discovery has fundamentally shifted: exploration is abundant, but physical verification governs the pace of breakthroughs.
Technology
LORE: Scaling physical verification with computation.
LORE is Lymeric’s proprietary search and verification engine. It expands physical verification with computation and converts a broad hypothesis space into a small number of experimentally defensible material assets.
Operating as a multi-fidelity verification funnel, LORE cascades from large-scale autonomous agent literature mining and combinatorial generation (10⁶+ candidates), through fast AI/ML surrogate screening (10⁴ candidates) and MLIP/molecular dynamics stability evaluations (10³ candidates), to high-precision ab initio DFT calculations (10² candidates). Only the most promising candidates advance to in-house pilot synthesis, electrochemical characterization, and external industrial partner validation—culminating in patentable composition-of-matter and process IP.
Product
Battery materials today, physical sciences tomorrow.
Our final product is not software seats—it is verified, high-value material IP assets.
Our first commercial validation domain is advanced battery materials—an industry where performance bottlenecks, supply chain constraints, and manufacturing costs create immense demand for breakthrough compositions.
We build and own defensible material IP portfolios, partnering with cell manufacturers, chemical producers, and OEMs through joint IP development and licensing agreements. Long-term, LORE scales across catalysts, functional ceramics, and broad physical science domains.
High Energy & Durability
Cathode Active Materials
Exploring novel composition spaces, gradient architectures, and stabilization doping to maximize usable energy density while preventing voltage fade and oxygen release.
Safety & Fast Ion Conduction
Solid-State Electrolytes & Interfaces
Computational search and verification of chemically stable solid electrolytes and protective interfacial coatings with high ionic conductivity and wide electrochemical windows.
Long-Term Frontier
Broad Physical Sciences Expansion
Extending the LORE exploration-verification engine to heterogeneous catalysts, functional electronic materials, and next-generation energy conversion systems.
Discovery-to-Asset Model
Lymeric operates as an asset developer and discovery partner. We capture economic value by owning composition-of-matter and process IP created through our proprietary discovery engine.
Team
Built across AI, chemistry, physics, and material science.
Core team
Experimental chemistry, biochemistry, and R&D process design; translating technical research problems into discovery pipelines, IP strategy, and commercial partnerships.
Ph.D. Chemistry, Columbia University; B.S. Chemistry, KAIST; former IQVIA consultant and National Cancer Center researcher.
Large-scale AI models, reinforcement learning, graph neural networks, and physics-based modeling for structured scientific and materials data.
Ph.D. Physics, UC Berkeley; B.A. Physics/Math, Cornell University; former chief scientist at Bone, with research experience at CERN and Oxford.
Computational chemistry, multi-fidelity physics simulations, MLIP/DFT workflows, and AI-enabled materials discovery.
Ph.D. Chemistry, Princeton University; B.S. Chemistry, HKUST; former CTO at TMAI and post-doc researcher at UNIST.
Contributors and advisors
Contributors
AI scientists, computational chemists, and materials engineers collaborating on candidate screening, simulation workflows, and experiment design.
Advisory network
Scientific and industry advisors across artificial intelligence, quantum chemistry, advanced battery technology, and intellectual property strategy.
Contact
Collaborate on verified material discovery and IP co-development
We partner with battery manufacturers, chemical producers, and strategic investors to co-discover, validate, and commercialize breakthrough material IP assets. Reach out to discuss pilot validation programs, discovery partnerships, or investment opportunities.
contact@lymeric.ai