Symbolic Reasoning, Multi-Physics Simulation & Agentic Intelligence
Symbol Science builds autonomous, zero-hallucination agentic workbenches that unify LLM reasoning, symbolic computation, and multi-physics simulation tools for engineering R&D.
Our Mission
"Real-world engineering discovery cannot rely on pure LLM text generation. It requires combining symbolic mathematics, multi-physics simulation software (CFD, FEA, FreeFem++, OpenFOAM), and automated visualization tools. We build the deterministic agent orchestration layer that turns complex engineering hypotheses into verifiable R&D execution."
Founder & Technical Direction
Yulin (Julian) Shi · 石玉麟
Founder & Chief Architect
Yulin (Julian) Shi holds a Master of Engineering from McGill University (specializing in structural dynamics & numerical simulation) and a Bachelor of Engineering in Aircraft Power Engineering from Beihang University (BUAA). With years of experience as a Senior Software Engineer at Alibaba (Lazada) and Booking.com—specializing in Domain-Driven Design (DDD), microservices, and workflow engine architectures—Julian leads Symbol Science in combining multi-physics simulation tools, symbolic reasoning, and LLM agentic workbenches for engineering R&D.
Our Core Pillars
Symbolic & LLM Reasoning
Combining LLM reasoning with symbolic math and formal logic to eliminate hallucinations in engineering & scientific workflows.
Simulation & Visualization
Automating multi-physics simulation (CFD/FEA), parameter sweeps, and high-dimensional plotting via code-generating agents.
Deterministic Agent Harness
Testable, modular tool orchestration connecting literature extraction, simulation execution, and automated report generation.
Partner or Collaborate
Whether you are an industrial R&D lab, an engineering team, or an investor, we welcome collaboration on scientific agent workflows.