Resume
Current Focus
I am currently working as a Software Engineer at Human Resocia Co., Ltd. and
assigned to Pacific Consultants Co., Ltd. in Tokyo, where I build enterprise AI
applications for infrastructure and government research initiatives.
Summary
Software engineer with experience building AI applications, LLM pipelines, RAG
systems, backend services, and full-stack products. My work focuses on semantic
search, knowledge retrieval, local LLM workflows, document analysis, and reliable
cloud deployment using Python, TypeScript, React, Node.js, PostgreSQL, Docker, and
GitHub Actions.
Areas of Expertise
| Languages |
Python, TypeScript, JavaScript, SQL, Bash, Java |
| Backend & APIs |
Node.js, Express.js, Flask, FastAPI, Spring Boot |
| AI / LLM |
LangChain, OpenAI API, Hugging Face Transformers, PyTorch, TensorFlow, scikit-learn |
| LLM Infrastructure |
Vector databases, Pinecone, FAISS, embeddings, semantic search, RAG |
| Frontend & Data |
React, Next.js, pandas, NumPy, Matplotlib, Jupyter Notebook |
| DevOps & Cloud |
Docker, Git, GitHub, GitHub Actions, DigitalOcean, Azure |
| Databases |
PostgreSQL, MongoDB, Redis |
| Languages Spoken |
English, Japanese, Tamil, Hindi |
Selected Projects
-
Enterprise RAG AI Assistant: Built an enterprise
Retrieval-Augmented Generation assistant for 400,000+ construction documents with
search, summarization, Q&A, PostgreSQL vector search, Docker, and GitHub
Actions.
-
AI Scenario-Based Presentation Generator: Developed a
conversational AI system that generates structured PowerPoint and Word documents
using RAG workflows, pgvector, Azure, and AI agents.
-
Local LLM Document Analysis & Evaluation System: Built local
LLM pipelines with OCR, embedding models, and GPU inference for discrepancy
detection, tender analysis, and proposal scoring.
-
Cross-Modal Translation Based on Triangle GAN: Researched and
implemented a deep learning model for cross-modal translation with strong semantic
preservation.
-
Code Mixed Sentiment Analysis: Built an ensemble sentiment
analysis model for code-mixed text using self-attention and XLM-R embeddings.
Publication
-
Code Mixed Sentiment Analysis (2024), presented at the
International Conference on Role of Artificial Intelligence and IoT in
Engineering, Technology and Science (ICRAETS'24).
Certifications
- Microsoft Certified: Azure Fundamentals
- Microsoft Certified: Azure AI Fundamentals
- Microsoft Certified: Fabric Analytics Engineer Associate
- Coursera - Stanford University: Algorithms Specialization
- Coursera - Google: Data Analytics Specialization
- Coursera - Google: Cybersecurity Specialization
- Coursera - Google: UX Design Specialization
- Coursera - IBM: Machine Learning Specialization
- Coursera - DeepLearning.AI: TensorFlow Developer Specialization
- Coursera - University of Michigan: Python Data Structures