Portfolio 2025

Nirai Hayakawa

> AI Researcher & Developer

18 years old

Bridging the gap between academic research and practical application. Specializing in Computer Vision, LLMs, and Game AI at CU Boulder.

View Projects Contact Me

Based in

Miyazaki, Japan

--:--

Core Stack

Python PyTorch TensorFlow OpenCV Rust TypeScript Astro AWS Docker Git

Connect

Email LinkedIn Download CV

About

Hello! I'm a researcher and developer with extensive experience in AI/ML applications for esports and gaming. I have a demonstrated ability to develop innovative solutions that bridge academic research with practical applications.

Currently, I am conducting research on LLM-driven algorithm optimization frameworks for NP-hard problems and have published research on computer vision and video analysis.

Tech Stack

Python C++ Rust TypeScript Swift Deep Learning Machine Learning LLMs Computer Vision Unity Unreal Engine AWS GCP Next.js Astro

Experience

Software Engineer Intern

2026/11-present

Not A Hotel Inc.

Recently working on backend development for "Kevin," the AI concierge at NOT A HOTEL, including the shared API platform used across the company.
Researching and implementing standard methodologies for AI agent evaluation frameworks.
Focusing on LLM as a Judge approaches for automated AI system assessment.
Designing evaluation infrastructure that can be reused across multiple AI-driven product domains.

R&D Product Developer Intern

2025/08/01-2025/10/31

CLINKS Corporation

Independently developed a comprehensive Proof of Concept from initial design to implementation.
Developed a speech-to-speech LLM agent system.
Prepared the demo for public release and stakeholder review.

Software Developer (Part-time)

2024/04/01-2025/06/31

Bird Fab Studio

Developed an automated texture packaging system using Python and LLM integration.
Reduced review failures to nearly zero and implemented automatic proofreading and translation.
Created a documentation system for the texture platform.
Built additional internal tools, including native applications.

Featured Projects

VALORANT Round Outcome Prediction

Collaborated with NAIST to build a transformer-based approach for predicting VALORANT round outcomes from minimap video analysis.

Python Deep Learning Computer Vision TimeSformer OpenCV

Texture Platform Automation

Built an automated texture packaging workflow for Unity and Unreal Engine with Python and LLM-assisted QA.

Python LLM Automation Productivity Tool

LLM Bias in Algorithm Evolution

Researching why LLMs drift toward conservative local optimization inside heuristic-search and algorithm-evolution frameworks.

LLM Research Algorithms Heuristics Bias Analysis

Speech-to-Speech LLM Agent

Designed and built a proof-of-concept voice agent that supports natural speech-in and speech-out interaction.

Python LLM Speech-to-Speech AI Agent R&D

AI Agent Evaluation Framework

Researching and implementing practical evaluation methodology for AI agents with an LLM-as-a-judge approach.

Python LLM AI Evaluation Research Automation

View All Projects

Latest Writing

March 8, 2026

Building a Scalable Backend for an LLM Chatbot During a Two-Week Internship at GMO Internet Group

During a two-week internship in GMO Internet Group's ML/WEB course, I joined a five-person team building an MVP and took charge of the backend for an LLM chatbot. This is a retrospective on the requirements, technology choices, design decisions, failures, and lessons learned.

internship backend llm typescript trpc

View All Writing