Profile avatar
nerddis.co
dx engineer @runpod.io; building → @wandler.ai
391 posts 352 followers 83 following
Prolific Poster
Conversation Starter

let's watch @andorofficial.bsky.social s02e01 one year later

the day has come i was finally able to move my runpod-worker-comfy from @blib_la to github.com/runpod-worke... in order to provide the support it needs to be the best @ComfyUI serverless endpoint on @runpod_io a lot of issues ahead, lets get them all solved

new video dropping soon: replacing my job at @runpod.io with agents: can mastra.ai really do it?

and just like that, i'm back using picthing...

am i building agents now?

having fun finding the perfect setup for vllm on @runpod.io endpoints to increase tokens per second

dropping tomorrow

vibe cooking @wandler.ai using @windsurfai.bsky.social while listening to mozart & friends LIVE → twitch.tv/nerddisco

useChat is almost done in @wandler.ai

after pushing this forward @runpod.io , we finally have an automated process to get updates from @github.com over to @docker.com hub into our website for our official pod templates. i'm very proud of the team and are looking forward to update all official templates next 🫡

democratize ai = free + private + local + open source access to ai for everyone last thursday i introduced wandler at the DataStax Berlin AI meetup. my talk, "democratize ai with the browser," showcased how to break down the barrier between web devs and ai running in the browser with js.

Made a gallery of such #SVG gradient + #filter effects blobs on @codepen.io: codepen.io/thebabydino/... Also includes an explanation of the how behind and interaction: the focal point of the `radialGradient` the blabs start from follows the cursor on hover. Different blobs for light/ dark mode. 😼

"prompt engineering" is just fancy copy-pasting at this point people tweaking prompts like they're adjusting a car mirror, thinking it'll make them drive better you’re optimizing nothing, you’re just guessing

Hello everybody! I launched a free newsletter where we deep dive into some of the toughest challenges for AI in production—especially now as POCs are everywhere, but real-world scalability gets the short end of the stick. I’d love for you to check it out! mlvanguards.substack.com

gen-commerce

what if ai just works?

the feeling when updating cursor fails...

thx @vite.dev for `?worker&inline`

claude-3.7-sonnet is really nice

me: hey cursor agent running on claude-3.7-sonnet, what's up? agent: hey, let's create a new chat app agent: let's create our own testing framework ... 1000 steps later ... agent: and this is how you would build your own browser

lfg cursor + claude-3.7-sonnet

vibe coding in cursor and out of the sudden it tries to run git log --since="2024-03-20" --pretty=format:"%h %s" | cat to find my recent commits and ended up in a total clusterfuck called cat /l、 (゚、 。7 l、゙~ヽ じしf_, )ノ

starting the day by generating a new user story using o3, so that my agent can implement it

took me 3 whole days as i had to refactor almost everything while testing with "webgpu + q4f16" and "cpu with wasm + q4" and worker vs main thread totally worth it, will do it any time again

gm

end the day with altered carbon

playwright is actually nice. have not used it intensively like this before.

the @wandler.ai loadModel e2e test will be finally completed soon.

we open sourced everything

it’s alive!

will release the site for @wandler.ai very soon...