🇯🇵 README_ja.md

I’m a backend engineer with strengths in data analysis and AI applications. I have experience in developing systems utilizing natural language processing, generative AI, and cloud services, and have driven projects both independently and within teams. Currently, I’m seeking a position where I can contribute to improving team technical capabilities and enhancing product value, while also gaining management experience.

💻 GitHub Stats

📈 Competitive Programming

• Category: Business System
• Responsibilities: Design, Coding, Testing, Data Analysis
• Role: Backend Engineer
• Technologies:

Developed a PoC tool that generates accurate kana readings from medical research texts for use in text-to-speech systems.
To handle domain-specific terminology and complex word variations in medical documents, I analyzed sample data and referenced overseas Japanese linguistic research resources to identify phonetic patterns.
Built the tool in Python, integrating MeCab with custom dictionaries and rule-based logic.
Achieved 99% reading accuracy, successfully completing the PoC.

• Category: Web Service / In-house Product
• Responsibilities: Design, Coding, Testing, Operation & Maintenance, Data Analysis
• Role: Backend Engineer
• Technologies:

Developed a generative AI system that provides precise and fast answers to user queries based on document data.
Implemented RAG (Retrieval-Augmented Generation), which significantly improved response accuracy and contextual relevance.

• Category: Web Service / In-house Product
• Responsibilities: Design, Coding
• Role: Backend Engineer
• Technologies:

Developed a system designed to generate high-quality images based on user-provided text using StableDiffusion.
As a backend engineer, I was responsible for managing the model invocation and generation workflow.
Additionally, implemented an AI-based text validation module to enhance the overall quality of generated images.

Extracts negative content from X posts and uses AI to generate ideas to address the complaints.

Generates illustrated diary entries based on posts retrieved from X.

Extracts common keywords from X posts to identify shared interests between users.