Tech Blog

Sharing AI transformation experience, industrial intelligence practices, and naval engineering insights

πŸ“… December 20, 2025 ⏱️ 15 min read

From 37-Year Marine Engineer to AI Model Evaluation Expert:
My Career Transformation Journey

A 59-year-old professor-level marine engineer shares his journey from traditional engineering to AI model evaluation. Discover how 37 years of engineering experience applies to LLM evaluation, Red Teaming, and industrial AI safety alignment. Learn the transformation path, ShipTechAI development story, and practical advice for engineers.

#AI Transformation #Career Change #LLM Evaluation #Red Teaming
Read More β†’
πŸ“… December 20, 2025 ⏱️ 12 min read

AI Agent in High-Speed Propeller Design:
ShipTechAI Case Study

In-depth analysis of using Deepseek and Claude collaboration to develop ShipTechAI, improving propeller design efficiency by 96%. Real cases: 3D interpolation algorithm implementation, intelligent diagnostic system development, dual-AI collaboration workflow. Complete technical path from traditional 4-6 hours to AI-assisted 10 minutes.

#AI Agent #Propeller Design #ShipTechAI #Engineering Application
Read More β†’
πŸ“… December 20, 2025 ⏱️ 14 min read

From Zero to Python Developer at 59:
My 10-Day Journey

Real documentation of the 10-day learning journey from complete programming beginner to independently developing ShipTechAI. Sharing the right methods for AI-assisted learning, engineer's advantages, actual development cases (data reading, 3D interpolation, GUI interface), and specific advice and learning roadmap for zero-base learners.

#Python Learning #Zero to Programming #AI-Assisted Learning #Engineer Transformation
Read More β†’
πŸ“… December 20, 2025 ⏱️ 16 min read

Deepseek vs Claude:
Real Engineering Comparison

AI model evaluation expert's real comparison based on ShipTechAI project: multi-dimensional evaluation of code generation speed & quality, architecture design capability, error diagnosis accuracy, professional knowledge depth. Detailed collaboration strategies, dual-AI workflow, prompt optimization techniques, and comprehensive scoring (Deepseek 8.4/10, Claude 8.7/10).

#AI Comparison #Deepseek #Claude #Model Evaluation
Read More β†’