Alessio Franceschelli

Principal Software Engineer ¦ .NET lover

[Alessio Franceschelli @AleFranz]

You drop an AI agent into your real codebase and it produces garbage. Wrong style, subtle bugs, side effects everywhere. The problem isn't the model. Your codebase isn't ready for AI. Think of it like handing off a biathlon relay with no map of the course 🥇🧵👇

[Alessio Franceschelli @AleFranz]

@antirez @anemll I wonder how feasible it would be to use it as a base to stream from RAM for non uniform memory systems (to always do inference on GPU)

[Alessio Franceschelli @AleFranz]

@nicolaibaaring @github @code I would say more between 5x-100x based on kind of task

[Alessio Franceschelli @AleFranz]

@konradkokosa congratz!

[Alessio Franceschelli @AleFranz]

@shanselman is the start menu CUDA accelerated now? 😝

[Alessio Franceschelli @AleFranz]

Full write-up here:

https://t.co/Nu5QkxVi3n

[Alessio Franceschelli @AleFranz]

A few days ago I published Burla. This follow-up is about the part I care more about: the design bets behind it. Toy examples are where mocking libraries look tidy. Real migrations are where they tell the truth. Thread below. 🧵👇

[Alessio Franceschelli @AleFranz]

I do not think the AI angle is marketing fluff.

Inconsistent APIs tax humans and LLMs in the same way: more context spent on framework quirks, less energy left for the code under test.

[Alessio Franceschelli @AleFranz]

@antirez are your tweaks all Metal specific or could this also be a good starting point for optimising a 128gb system with hybrid CPU/CUDA inference via layer offloading?

[Alessio Franceschelli @AleFranz]

Repo, NuGet, and the broader story are in the blog post:
https://t.co/9lCkjAIn36

[Alessio Franceschelli @AleFranz]

I finally published Burla, a .NET mocking library I have been rewriting for months while pressure-testing models, local setups, and agentic coding workflows. I needed a real project, not one more opinion about AI. Built with LLMs in the loop, for the LLM era. Thread below. 🧵👇

[Alessio Franceschelli @AleFranz]

Here are the Docs:
https://t.co/ZjK060n98m

[Alessio Franceschelli @AleFranz]

The future winner will not just be the model that is slightly better.

It will be the model that gets the job done reliably, quickly, and cheaply enough to become the default.

Full post: https://t.co/wXRq0L6JSF

[Alessio Franceschelli @AleFranz]

If you're still defaulting to Claude Opus in GitHub Copilot, you're probably paying a hype tax. 🧵👇

Opus 4.6 = 3x requests.
Opus 4.7 = 7.5x.
GPT-5.4 = 1x, even with xhigh reasoning in VS Code.

I do not see a quality gap remotely close to that.

[Alessio Franceschelli @AleFranz]

GPT-5.4 also tends to be faster and less rate limited.

People underestimate what slow loops do to the workflow: fewer iterations, worse focus, and less willingness to explore alternatives.