Skip to content

DEV Community

prodbymarcu

Audio Engineer - Producer - Songwriter e/acc Local AI is the future x: @marcvgarcia

Florida Joined on Oct 1, 2026

Oct 23

Your camera's wrong password error can be three defaults deep

#discuss #community #productivity #selfhosted

4 min read

Oct 22

A 31M TTS model ran 2.6x faster than the 2.37B incumbent

#discuss #community #productivity #ai

4 min read

Oct 22

A 100% progress bar on a 250 GB download proves nothing

#discuss #community #productivity #llm

4 min read

Oct 21

Three layers said healthy while the engine slept 104 minutes

#discuss #community #productivity #debugging

5 min read

Oct 20

Your smart bulb ate the color and lied about it

#discuss #community #productivity #hardware

4 min read

Oct 19

A half-million-token prompt found both needles on my 3090

#discuss #career #community #ai

3 min read

Oct 19

Why does your voice agent go silent mid-call?

#discuss #productivity #community #ai

5 min read

Oct 18

I asked an LLM to spec my dual-GPU power plan, and the second supply could never start

#hardware #automation #diy #learning

5 min read

Oct 18

My art director agent scored its own render 9/10, then frame stepping called it 4/10

#ai #llm #gpu #video

5 min read

Oct 17

My self-hosted ad blocker poisoned Apple's push servers and killed iMessage

#dns #networking #docker #selfhosted

4 min read

Oct 17

I almost rebuilt a project for the wrong reason, a 20 minute audit caught it

#productivity #security #softwaredevelopment

4 min read

Oct 17

I built a handheld mmWave presence radar, and the first bug was a board that never ran my code

#esp32 #embedded #arduino #hardware

4 min read

Oct 16

I made my 3090 single-lane for agent turns, and the first live duel caught my lock file at the wrong path

#ai #llm #devops #concurrency

4 min read

Oct 16

I put the 381 tok/s Qwen stack on one RTX 3090, real chat does 118 and the headline mode crashes

#llm #inference #gpu #docker

5 min read

Oct 16

I read a CUDA inference engine's symbols with Ghidra, then made it 12% faster on my 3090

#python #cuda #llm #inference

5 min read

Oct 16

I'm chasing one of a hundred free crab robots, and the contest rules decided the shape of the work

#python #robotics #simulation #physics

4 min read

Oct 16

I built an auction engine that refuses to bid for me, and the platform terms say why

#python #sqlite #testing #automation

4 min read

Oct 16

I replaced NFS with a FUSE filesystem on my homelab, and macOS ate three working days

#docker #macos #filesystem #networking

4 min read

Oct 1

I benchmarked Cloudflare's new open decision model against the hosted API it's trying to replace

#machinelearning #ai #cloudflare #llm

4 min read

Oct 1

My local LLM fixed a bug overnight. The "success" was the bug.

#ai #llm #python #automation

4 min read

loading...