AI
AI Hallucinations: Fixing "Nice School Code" Errors with an API
My AI was hallucinating "Nice School Codes," but I fixed it using the official open API. I'm sharing the API integration process and my know-how for solving this problem.
AI
Improving AI Content Demand Prediction Accuracy: Enhancements Through Integrating Real-World Data
This post details the process of integrating real-world data to improve the performance of an AI content demand prediction engine. Learn how to leverage Naver Data Lab and YouTube view count data.
AI
Vertex AI Grounding Cost Gap: Diagnosing the Missing $1300 on My Solo VM
I discovered a significant cost discrepancy in my Vertex AI usage, realizing that token counts alone don't reveal the full picture of LLM expenses. Grounding costs were hidden until I implemented explicit logging.
AI
Gemini Model Management: Ending Inefficiency! The Secret to 3x Faster Cost Tracking with Model Registry
Sharing my experience solving the complexities of Gemini model management and the issue of inefficient cost tracking using Model Registry.
AI
4 Pitfalls Discovered After Migrating from Anthropic to Gemini
I switched from Claude Sonnet 4.6 to Gemini 2.5 Flash to save 80% on costs, and encountered these real-world bugs. I hope you can avoid the same mistakes.