Serene Mathew

Engineering @babbel. Former @marksandspencer / @tcs

[Serene Mathew @serene_matt]

@sebastienlorber React Native abandoned years ago – https://t.co/Knhq7s45yM

[Serene Mathew @serene_matt]

We are not buying anyone else’s factory. We are building our own verification system and wanted to give it a name and the term “factory” got stuck.
Anyway, I am on factory mode for the next two weeks with AI Heroes cohort.

[Matt Pocock @mattpocockuk]

Just wanted to defend the term "Software Factory" for a second.

First, the caveats:

It's bullshit. It sounds fancier than it really is.

It's a bandwagon. People are using it to get you and their VC's hyped.

But it's also referring to a real change.

Earlier this year, most SOTA teams were using agents to work on the product. Shipping features, fixing bugs.

But now, they're building automations to help agents work on the product for them. They've moved up a layer.

What do we call the thing they're building? Loops? Workflows? Some other generic label?

"Software Factory" is as good a term as any other.

- "We're not working on the product, we're working on the factory that builds the product."
- "We're not product engineers, we're factory engineers."

I like it. I'll continue using it.

[Serene Mathew @serene_matt]

@addyosmani Hope the teach style remains. It's fun and almost felt illegal to write the code.

[Serene Mathew @serene_matt]

@propresenter loading with “Leaving the 99 to find one missing slide …” 🤣

[Serene Mathew @serene_matt]

@Metaliberal @mattpocockuk Thinking of adding a limit to the number of lines of code for a function, otherwise split it. seeing large chunks of code in one module.

[Serene Mathew @serene_matt]

@orcdev I agree ☝️ New dopamine

[Serene Mathew @serene_matt]

@sebastienlorber Compute concepts

[Serene Mathew @serene_matt]

Still getting a lot of "Your fix is better than my fix" from agents ... smirk

[Serene Mathew @serene_matt]

@burkelibbey Rewriting a monorepo to a totally different stack is not the easiest thing with agents. I ended up having double agent instructions

[Serene Mathew @serene_matt]

Keeping an eye on tsrs👀

[Serene Mathew @serene_matt]

@theo Shift-left testing

[Serene Mathew @serene_matt]

@emilkowalski What’s your take on visual regression in CI after all these verifications? I find it less useful now that agents are the ones watching the progress.

[Serene Mathew @serene_matt]

@theo I do use the terminal drawer in t3Code though

[Serene Mathew @serene_matt]

@TechCrunch Exactly what the really younger generation needs. I have seen them struggle hard when switching from a touch device to a laptop.

[Serene Mathew @serene_matt]

@mattpocockuk Already banned. Curious if you would put this in the implementer agent or the reviewer.

[Serene Mathew @serene_matt]

@RhysSullivan Fighting with whom? agent or human? Spent a good amount of time today going back and forth on picking a build tool

[Serene Mathew @serene_matt]

Got to refactor the software factory it seems

[Addy Osmani @addyosmani]

Before adding another subagent ask what problem it's solving. A bloated system prompt is often the issue.

Skills and tool search can fix that without splitting context across agents. My rule of thumb is to keep a subagent when the work is parallel, needs fresh context or one agent really can't pick the right tools.

Otherwise, fold it into the main loop, turn its prompt into a skill, run your evals and keep whatever wins.

Good write-up:
https://t.co/1gcrNNpQru

[Serene Mathew @serene_matt]

@emilkowalski Love the animation-performance skill! Used to have these instructions scattered across CODING_STANDARDS.md and prompts.

[Serene Mathew @serene_matt]

There’s no easy way to search a previous conversation with Grok @bot other than letting it search and use some more tokens, is there?

[Serene Mathew @serene_matt]

@orcdev Seeing the threads auto-settle based on PRs merged is so fun.