Computer Science Student | Software Engineer | Agentic Software Engineer

Building software that turns human intent into connected, reviewable actions through LLMs, agents, tools, and services.

📍 Bangkok, Thailand  •  📫 anothai.0978452316@gmail.com  •  🌐 taichi112.works

I am a Computer Science student developing toward Software Engineering and Agentic Software Engineering through project-based practice.

My direction is to build software products where people can express what they need in natural language, while LLM-powered agents coordinate with tools, services, data, and other software systems to help complete the workflow. For important actions, I value human-in-the-loop control so users can review, choose, and approve outcomes before execution.

I am interested in the foundations of AI, Machine Learning, Deep Learning, and Data, and I continue learning them to better understand the intelligent capabilities used in modern systems. My primary engineering focus is applying those capabilities inside reliable software products — supported by maintainable architecture, scalability, clean design, and practical user value.

I want to develop software that moves beyond rigid, manual interaction flows. Instead of requiring users to understand every configuration option or repeatedly navigate multiple systems, an agentic product should help interpret the user's goal, coordinate the necessary tools, and present safe actions for approval.

Human Goal in Natural Language
            ↓
Intent Understanding & Context
            ↓
LLM / AI Agent Workflow
            ↓
Connected Tools, APIs, Data & Software Services
            ↓
Recommendation or Prepared Action
            ↓
Human Review / Approval (HITL)
            ↓
Controlled Execution

Examples of experiences I want to build include:

• A user describes a deployment requirement, and the system prepares suitable infrastructure configurations for review.
• A user describes what they want to buy, and the system finds matching options, compares trade-offs, and prepares a selection for confirmation.
• An organization connects operational tools, documents, and services so an agent can assist with repetitive workflows while preserving human control.

Agentic workflow · Python · MCP · Google Calendar API

Built a Python-based AI agent that interprets natural-language scheduling requests and integrates with Google Calendar operations, including appointment overlap detection. This project represents my interest in software agents that connect user intent with real tools and controlled actions.

Software product architecture · TypeScript · Structured data · Personalized workflows

Designed a structured system for managing skills, projects, and experiences so users can produce role-specific portfolio or resume versions. The system creates a foundation for future LLM-assisted recommendations and human-reviewed document generation.

Applied intelligent workflow · Python · Image Processing · OCR · HITL

Engineered a pipeline for extracting and translating text from manga images using image processing and OCR, with a human review flow for translation quality and contextual accuracy.

Developer tooling · CLI · Extensible system design

Built a command-line tool for automating project bootstrapping and reusable development workflows. Its design is intended to evolve toward LLM-assisted software setup through natural-language requirements.

Software engineering foundation · TypeScript · Next.js · GoF Patterns

Created a practical system for applying software engineering concepts, including GoF design patterns and maintainable component architecture. The implementation uses a browser-based interface, while the project's purpose is architectural learning and maintainable software design.

Intelligent interaction · AI-assisted content · Real-world participation

Developed a platform that connects physical and digital event participation, applying AI-assisted question generation to support more meaningful professional interactions.

• Designing agentic software systems where LLMs and AI agents support real product workflows.
• Exploring Model Context Protocol (MCP) for connecting agents with tools, data sources, APIs, and software services.
• Learning LangChain and LangGraph for tool-using workflows, orchestration, persistence, and human-in-the-loop control.
• Continuing to build foundations in AI, ML, DL, and Data to better evaluate and apply intelligent capabilities responsibly.
• Applying clean architecture, SOLID principles, design patterns, scalability, and maintainability to AI-integrated software.
• Building systems that let users interact naturally while keeping high-impact actions transparent and reviewable.

Currently learning and exploring: LangChain · LangGraph · MCP-based integrations
LLM tools used in development: Gemini · Claude · GPT

I am seeking opportunities to contribute as a Software Engineer Intern while growing toward Agentic Software Engineering: designing reliable software products where LLMs and AI agents help connect human goals with tools, data, and services through natural-language interaction and human-approved execution.

Areas of Interest: Software Engineering · Agentic Software Systems · Intelligent Product Development · LLM Integration · Tool-Connected Workflows · Human-in-the-Loop Systems