Francesco Pochetti

If you have trained a model, you want to deploy it to SageMaker, and your first instinct is to start screaming, please😌 and keep reading. Today, I am launching a course: How to deploy an ML model to SageMaker with a custom Docker container 👉 Get it here: frapochetti.gumroad.com/l/mdpbvi

@thealokverse @Yuchenj_UW Got demoted by Gemini

@Yuchenj_UW That's a great question. (chilling in a jacuzzi, jk)

Sergey used Gemini in a surprisingly creative way. He asked it inside an internal Google chat, “who should be promoted in this chat space?” AI picked up a young woman engineer who is not vocal, and she was promoted. AI can reveal value that human systems overlook. AI can help CEOs of Big Corps spot true talent rather than people who are just loud or political, and it's probably more objective than the middle managers. How far are we from an AI CEO? Sergey Brin in founder mode using AI to cut through bureaucracy is just fascinating.

The full story covers: disasters, pivots, technical architecture, and what actually works when vibe-coding with LLMs. Tech: Next.js 15 • React 19 • Prisma • Replicate • AWS Bedrock #AI #LLMs #VibeCoding #BuildInPublic

3 complete pivots: Feb: AWS Amplify (15-min builds, mysterious failures) Apr: Vercel + Supabase (too many platforms, blind LLM debugging) Aug: EC2 + local dev → shipped The breakthrough: KISS. No S3, no complexity. Just Next.js + SQLite + nginx.

I finally shipped Visual Neurons: a personal AI studio for image/video generation. 9 models. One interface. Zero lines of code written by me. Claude Sonnet 4.5 + Cursor built it. 📖 Full story: francescopochetti.com/building-visua… 🎥 Demo: youtu.be/1TS7quWMmIY Here's what it took 🧵

Working on adding Runway Gen-4 video generation. Close but not there yet. runwayml.com/research/intro…

🎉🎉 Seems I got Gemini Image Edits to work! 🎉 🎉

Getting close to Chat with Gemini integration

@richdevelops Yes!

When I thought I was done with outpainting, I started stumbling on frontend timeouts. The StabilityAI API call is successful though (I see it in 💸). Lambda logs confirm too. Seems AppSync plays tricks on me. Here's how ChatGPT puts it 👇 (and yes, I have to move to async)

Stability's outpainting is not bad at all. 300px completely "invented" in all directions.

What happens in that scenario? Stability finishes the job in, say, 40 seconds, returns HTTP 200, and you get billed. However, AppSync has already cut off your resolver at 30 seconds, so your front end sees a “timeout” or “network error,” rather than the final 200 result. ```

``` If you are calling the Stability outpainting API from an AWS Lambda resolver (via AppSync), remember that: AppSync imposes a 30-s timeout for all resolver calls. Even though your Lambda function can be configured for a 120-s timeout, AppSync itself will cut off at 30 seconds.

Managed to add Outpainting from StabilityAI 🎉 Now hardcoding 100px in each direction. #tag/Edit/paths/~1v2beta~1stable-image~1edit~1outpaint/post" target="_blank" rel="nofollow noopener">platform.stability.ai/docs/api-refer…

Seems to work

I am close for Style Transfer from Stability. It'll happen tomorrow.

Integrated StabilityAI image generation. After the 20-commit-3-days refactor it was a breeze 🙌 It boiled down to implementing the `generateImage` func inside the `stabilityProvider.ts` and the dispatcher propagates to Lambda and the frontend. #diff-df55f3de0e8f2ebc323d18396135b14de6513930cad5376a39683a1bed6c8955" target="_blank" rel="nofollow noopener">github.com/FraPochetti/vi…

Claude is happy with the refactoring 🫡 github.com/FraPochetti/vi…

Weekend plans! Believe it or not but it's not fully refactored yet 😱

@robmarkcole @AnthropicAI `o3-mini-high` absolutely kills it. Claude 3.7 is very very good but it requires too much fiddling with the context. Super annoying. Gemini image understanding is top notch (together with GPT4 tbh). Never really given it a proper shot as a coding assistant

@Fra_Pochetti @AnthropicAI I hear even Gemini is worth a shot now

Not sure what they are doing at @AnthropicAI, but Claude is basically unusable for any reasonably long context. ChatGPT handles the same queries without any issues.