My Top 3 read articles for April 2025

1. TrapC: Can It Save C from Rust?

C has taken a beating lately for not being memory safe — and Rust is gaining traction precisely because of that. To make matters worse, CISA is now pushing software vendors to move away from memory-unsafe languages altogether. Major tech players have already started backing Rust, rewriting parts of their system software in it.

Against this backdrop, Robin Rowe’s proposal for the TrapC extension to C is potentially game-changing. According to the proposal, compiling C code with a TrapC-aware compiler automatically upgrades all pointers to Memory Safe Pointers, with built-in checks and enhanced error handling mechanisms

Rowe asserts TrapC is on track to be released in 2025. If done well this can totally kill the C to Rust migration industry. It will be interesting to see how this unfolds.

👉 Read the full post

2. Ollama: Run LLMs Locally, No GPU or API Keys Needed

Ollama makes running open-source LLMs locally dead simple — no cloud, no API keys, no GPU needed. Just one command (ollama run phi) and you're chatting with a model that lives entirely on your machine.

Built by a small team of ex-devtool and ML engineers at Ollama Inc., the project wraps the powerful but low-level llama.cpp engine in a smooth developer experience — handling model downloads, quantization, and inference behind the scenes.

Whether you’re hacking on a side project, building an AI CLI tool, or just exploring what local models can do, Ollama cuts through the setup pain and gets you straight to the fun part.

Think of it as Homebrew, but for brains — local-first, developer-friendly, and privacy-respecting. It just works.

Minimalism and simplicity are absolutely central to Ollama’s brand and product philosophy as by the design of their home page.

👉 Read the full post

3. Are AI-Generated Database Schemas Production-Ready?

AI-powered tools like ChatGPT, Gemini, and DeepSeek have changed how developers approach database design. Today, teams routinely use AI to generate schemas, validate ideas, and build quick prototypes. But how good are these AI-generated schemas in real-world applications?

As we experimented with AI for designing our PostgreSQL database, we noticed that AI often missed crucial best practices. In this post, we will share some real-world scenarios where AI could have done a better job, and we will also compare PostgreSQL and MySQL in terms of storage optimization and data type flexibility.

When we started using it, we were expecting AI to not only generate the schema quickly but also for ensure by default

• Implement the best practices or thumb rules of database design.

• Infer relationship & take care of suggesting appropriate constraints.

• Choose the correct data types for the column.

• Optimize from a storage and query performance.

The reality was quite different & in this post we present few scenarios where the team thought AI could have done a better job in the first go.

👉 Read the full post

I work with early-stage startups and growth-stage tech teams as a Fractional CTO & CISO, helping them move faster without cutting corners.

🔗Connect with me on LinkedIn to start a conversation.
🔔Follow me on Medium