Summary
Dong-ha Shin is a Senior Associate in AI Research specializing in on-device AI, computational optics, and real-time vision systems with eight years of hands-on experience bridging photonics research and edge deployment. He led the CES 2025 demo integrating STT–LLM–TTS on a Qualcomm entry-level SoC to achieve ~1.5s offline latency, and currently develops efficient vision models and NPU-specific optimizations for road-hazard detection at DareeSoft. His background spans holography, diffractive optics, and light-field simulation from research roles at KNU and POSTECH, where he built differentiable light-wave simulations and CGH optimization pipelines. Dong-ha holds a strong academic record in Electrical and Electronics Engineering (GPA 4.12/4.5) and has authored papers and patents on hologram upscaling, metasurfaces, and autonomous sensing innovations. Notably, his work combines experimental optical setups with production-ready edge ML—turning nanoscale device insights into deployable, latency-optimized inference stacks. He is mission-driven to make real-time imaging faster, safer, and more immersive and is actively open to collaboration.
8 years of coding experience
3 years of employment as a software developer
학사, Electrical and Electronics Engineering, 4.12/4.5, 학사, Electrical and Electronics Engineering, 4.12/4.5 at 경북대학교(Kyungpook National University)
English, Korean