Hikaru Ikuta is a research engineer based in Tokyo with a Ph.D. in Information Science and Technology from the University of Tokyo and 12 years of hands-on experience across research labs and industry. His background spans low-level systems, compiler infrastructure, and applied graphics—work that includes implementing backends for esoteric language VMs (elvm) and bootstrapping a Lisp in a boot sector. He has research internship experience at Microsoft and Adobe and has shipped prototypes and papers from internships and industrial research roles, bridging academic rigor with product-minded engineering. At Sony and multiple startups he applied deep learning and rendering techniques, including a SIGGRAPH ASIA–noted texture transfer and a shader for photorealistic hair rendering. Colleagues describe him as a practical innovator who moves between theory and implementation comfortably, from binary lambda calculus backends to real-world I/O fixes. He currently continues research-focused engineering at Mantra Inc., bringing compiler-level insight to applied problems.
12 years of coding experience
2 years of employment as a software developer
University of Tokyo
Bachelor of Engineering - BE, Applied Physics and Physico-Informatics, Bachelor of Engineering - BE, Applied Physics and Physico-Informatics at Keio University
Contributions:3 reviews, 28 commits, 8 PRs in 4 months
Contributions summary:Hikaru contributed to the bootstrapping of a Lisp interpreter. Their work involved fixing bugs related to the `EQ` function, variable referencing, and register initialization. They also implemented essential I/O capabilities, including the `READ` and `PRINT` functions, and made modifications to the test file. These changes expanded the interpreter's functionality and improved its stability.
Contributions:15 commits, 7 PRs, 24 comments in 1 year 6 months
Contributions summary:Hikaru contributed significantly to the `elvm` repository by implementing backends for various esoteric programming languages and compiler infrastructure. Their primary focus was adding support for the QFTASM (Conway's Game of Life) and Lazy K backends, which involved writing code for compilation and interpretation. Additionally, the user revised the QFTASM backend and implemented computed goto and a binary lambda calculus backend, which demonstrates a deep understanding of compiler design and target-specific code generation.
llvmcompilersbisoninfrastructurecompiler
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.