Thanit Kebsiri

Thanit Kebsiri

AI Engineer • Educator • Veteran Software Architect

25+ years leading enterprise software development, architecture, and teams (Java, SOA, microservices, integration). Since 2019: Deep focus on Generative AI, LLMs, RAG-optimized agents, AI + cybersecurity. Creating practical, high-impact online training for AI Engineering professionals (target: Master-level program launch 2026).

GitHub LinkedIn Medium Email

About Me

With more than two decades in IT, I have managed teams, consulted on enterprise solutions, and architected complex systems across insurance, finance, government, and more. Key strengths: SOA, ESB (Mule), BPM workflows, microservices, legacy integration, project leadership (PMP/PRINCE2), and full product lifecycle delivery.

From 2019 onward, I transitioned fully into AI research and education. I explore deep reinforcement learning, LLM integration/finetuning (Mistral 7B via Hugging Face), agentic tools (LangChain, GPT Engineer/Pilot, Open Interpreter), computer vision (OpenCV), and increasingly: RAG for context-aware systems, AI agents (inspired by spec-kit, SKILL.md), and cybersecurity in AI (secure prompting, vulnerability mitigation in agents).

Based in Bangkok, I freelance as an instructor and researcher — focused on bridging traditional software engineering wisdom with modern AI to create accessible, production-ready training materials. Goal for 2026: Launch a comprehensive online Master in AI Engineering program emphasizing practical agent development and secure AI workflows.

Core Skills & Expertise

AI & Generative AI

  • Python, PyTorch, TensorFlow, Hugging Face, scikit-learn, Keras
  • LLMs & Tools: LangChain, OpenAI API, GPT Engineer/Pilot, Open Interpreter
  • RAG optimization, LLM finetuning (Mistral 7B AutoTrain), Deep RL, OpenCV
  • AI Agents, prompt engineering, AI + cybersecurity (secure agents, pen-testing integration)

Software Architecture & Dev

  • Java/J2EE, Spring Boot, Struts, Hibernate, Microservices (Kubernetes, Docker)
  • Integration: Mule ESB, Camunda BPM, Spring Integration, Camel
  • SOA, legacy modernization, CI/CD, Git
  • Project & Risk Management: PMP, PRINCE2, Ms Project, Rational tools

Tools, Platforms & Education

  • Databases: Oracle, Sybase, MySQL, MSSQL, DB2
  • Security: Kali Linux pen-testing
  • IDEs: Jupyter, Eclipse, PowerBuilder, PowerDesigner
  • Training: Open-source Java, enterprise frameworks, beginner-to-advanced LLM/AI courses

Professional Experience

Freelance Software Development Instructor & AI Researcher

Jan 2019 – Present

  • Develop and deliver online training: LLMs for beginners, AI-assisted development, deep RL basics.
  • R&D: Integrate GPT-2 into microservice testing, OpenCV for car insurance claims, RAG/agent optimization (2025–2026 focus: spec-kit/SKILL.md style agents + AI cybersecurity).
  • Articles on effective prompting (Grok-3 masters, LLM coding pitfalls).

Software Development Manager / Architect – Asia Insurance 1950 Co.,Ltd

Nov 2011 – Sep 2018

  • Led architecture & implementation of next-gen Car Claim System: CRM integration, Mule-ESB across legacy systems, BPM workflows, mobile frameworks, surveyor GPS routing.
  • Trained teams on open-source solutions, integration patterns, best practices.
  • R&D: P2P video conference, BPM optimization.

Earlier Leadership & Consulting Roles

1995 – 2011

  • Sybase DBA Consultant (Siam Commercial New York Life), Software Architecture Manager (Coda/IBM – SOA for MOL Project), Senior Consultant (Ness, JSG – e-commerce/CRM/portals).
  • Software Developer/Project Manager (P.S.I Technology – e-commerce/Client-Server).
  • Instructor/Manager (Sybase Thailand – PowerBuilder, PowerDesigner, Java R&D).

Selected Projects & R&D

Generative-AI-for-Everyone

Hands-on, free course introducing generative AI concepts and tools.

GitHub →

Ai-Driven-Artifacts

Spring Boot microservice with Aider AI for dynamic online training support.

GitHub →

Next-Gen Car Claim System (Asia Insurance)

Architected integration framework (Mule-ESB + BPM) connecting legacy CRM, vendors, GPS routing, mobile comms.

2025–2026 AI Research Focus

RAG-enhanced LLMs, agentic systems (spec-kit, SKILL.md patterns), cybersecurity hardening for AI agents/training platforms. Foundation for upcoming Master in AI Engineering online course.

Active exploration via GitHub forks: Skill_Seekers, sc-agent-skills-files, anthropics-skills

Recent Writings

Education & Professional Training

Ramkhamhaeng University

Business Administration: Marketing (1992)

Key Professional Trainings

  • AI Deep Reinforcement Learning & Neural Networks
  • Microservice Architecture, Mule-ESB, Enterprise Integration Patterns
  • SOA Architect, Software Quality Assurance (IBM/Coda)
  • Professional Project Management (Sybase)
  • OOP with PowerBuilder/Java