I Pulled My Data From the Cloud: Do I Regret It?
Did I regret moving my data on-premise and breaking free from cloud dependency? I'll share the technical and operational reasons behind this decision from my.
333 posts
Deep dives into software, cloud computing, AI and technology trends.
Did I regret moving my data on-premise and breaking free from cloud dependency? I'll share the technical and operational reasons behind this decision from my.
After Plex Pass's pricing policy change, I detail my experience switching to Jellyfin, from setup to performance, security to user experience…
I'm sharing the challenges I faced and the lessons I learned when deciding to adopt new technology. On the risks of early adoption and correct timing…
I explain why Kubernetes isn't the only solution for every project, highlighting the advantages of simplicity and cost-effectiveness based on my 20 years of.
As AI model token costs rapidly increase, I explain how you can reduce your bill using practical methods I've experienced.
Learn how to build your own AI agent using Python, LangChain, and the OpenAI API. A step-by-step guide to automating tasks.
I ran my own AI agents autonomously for 6 months. In this process, I encountered successes, disappointments, technical details, and my cost analysis…
5 key reasons why Proxmox will strengthen your homelab in terms of high availability, storage, networking, and security.
Exploring the Microservice Communication Protocol (MCP) standard, which solves the incompatibility problem between AI models, using a USB-C analogy and my own.
I analyze the performance of different LLM models based on their workloads. Comparing GPT-5.5, Claude, Gemini, and DeepSeek to help you choose the right.
With 20 years of system architecture experience, I examine whether managing your own servers is a pleasure or an inevitable need.
I examine the cost increases brought by GitHub Copilot's new token-based pricing model and the strategies I've developed to counter it.
One of the most expensive lessons I've learned in my career: Unnecessary complexity always invites disaster. The power of simplicity and why it's critical…
With 20 years of field experience, I examine the fundamental differences, commonalities, and operational challenges of system architecture and AI solution.
I argue that vibe coding is outdated and has been replaced by Karpathy's 'Agentic Engineering' approach. This new era focuses on AI agents in engineering...
In 2026, we'll explore the differences, advantages, and disadvantages between AI coding tools like Cursor and Claude Code to help you make the right choice...
I examine the potential dangers of AI agents in production environments through a real data loss scenario. Why should we be careful?
With 20 years of experience, I share the promises and challenges I faced in social network development, from scale to security, moderation to sustainability.
Despite the dazzling promises of distributed systems, my 20 years of experience have often shown me the value of the simplicity and control that centralized.
With 20 years of systems architecture experience, I discuss why a single VPS is often sufficient and how adding more can be a waste of resources.
Local build cache or remote cache in your CI/CD pipelines? I dive deep into the balance of speed, cost, and efficiency.
As artificial intelligence rapidly enters our lives, I discuss the limits of AI and what it has yet to achieve, drawing on my 20 years of experience in system.
Mustafa Erbay's pragmatic take on whether using a vector database is truly necessary for your AI projects, exploring trade-offs and alternative approaches.
Advantages, disadvantages, and considerations for building your own push notification system instead of relying on Google Firebase Cloud Messaging (FCM) and.