Isam A

Founder & CEO of Core Industry

[Isam A @Ismbuilds]

@_sameerrr0 Left for sure, right is too much

[Isam A @Ismbuilds]

@koppxo Looks sick 🙏🏻

[Isam A @Ismbuilds]

@superjohnca @Abo3losh01 @alexandr_wang NPC?

[Isam A @Ismbuilds]

@ardent__dev Depends but mostly pnpm

[Isam A @Ismbuilds]

@SadAlbert10 This is a poor people issue. I feel u tho with my pro plan

[Isam A @Ismbuilds]

@aslambeg84 Cloudflare for sure

[Isam A @Ismbuilds]

@Surendar__05 Claude has been dogshit lately

[Isam A @Ismbuilds]

@md_kasif_uddin GPT everyday! Claude has been complete garbage the last couple of weeks and I’m clearly not the only one noticing. Anthropic keeps posting pretty charts, but the actual product is laggy, dumbed-down, and borderline unusable. User experience is straight dogshit right now.

[Isam A @Ismbuilds]

@awsdevelopers 01100010 01111001 01100101 00100000 01110111 01100101 01100101 01101011 01100101 01101110 01100100 00101110 00100000 01110100 01101001 01101101 01100101 00100000 01110100 01101111 00100000 01100010 01110101 01101001 01101100 01100100 00101110

[Isam A @Ismbuilds]

@xikhar I switched from Claude to Codex because I was spending 200$ on a plan I’d use up in two days. Codex lasts much longer.

[Isam A @Ismbuilds]

I keep separate notes for tools I test and tools I trust in production. The first list changes every week. The second changes slowly. New is useful, but repeatable is what earns a place in the system.

[Isam A @Ismbuilds]

I build software products and test AI tools every day. Some change the way I work. Others only look impressive in a demo. I am starting this thread to document the difference.

[Isam A @Ismbuilds]

I keep a simple test for new AI tools: can they remove one repeated step without creating three new checks? If not, the demo may be impressive, but the workflow is not better.

[Isam A @Ismbuilds]

The standard is simple. If a tool saves time, improves quality, or changes how I build, I will explain why. If it does not, I will say that too.

[Isam A @Ismbuilds]

I will share practical notes from building: prompts that hold up, LLM features that fail in real workflows, system decisions, useful tools, and the work between an idea and production.