CEO @smartsquadsrl · Founder @inksquadart. Building software with a UX/UI-first product mind.
Udine, Friuli-Venezia Giulia
Joined October 2011
- Tweets773
- Following349
- Followers183
- Likes3.6K
Your first screen should be where users already go, not where you hope they'll go.
We moved the AI chat to the front of Inksquad. Usage went up.
What's the first screen of your app, and is it the one people actually want?
@inksquadart
Empty marketing screens and product heroes get boring fast.. this keeps them alive without a heavy animation stack
iim.smartsquad.io
New open-source drop: Images in motion.
Continuous image lanes for the web; pure CSS, no JS frame loop. ~6.5 KB gzip. Works with @reactjs, @vuejs, plain JS; @expo / @NativeScript via WebView.
Try it → iim.smartsquad.io
Built this for real @expo apps that need Apple auth without three different flows.
If you’re on SDK 58 and still wiring SIWA by hand, this saves a week.
Supervised and shipped with our CTO @massimodeluisa
We open-sourced Sign in with Apple for @expo SDK 58.
One API + one Apple-style button on iOS, Android, and web. Same credential shape everywhere. Nonce hashing built in. Adapters for Supabase, Clerk, and Firebase.
Docs + install → exsi.smartsquad.io
This is the bit I've been waiting for.
An agent that answers with UI you can actually click and edit. Trip plans, comparisons, dashboards, right there in the reply.
OpenUI 1.0 makes agents feel like proper software. Really cool work.
For 30 years, technology has moved in one direction: turning information into intuitive interfaces
Somehow, the smartest software we’ve ever built took us back to a textbox
We’ve spent the last 2 years at @openuidev trying to fix that. 6 months ago we open-sourced it: 10K+ GitHub stars, 6M+ downloads
ChatGPT shipped Intelligent UI and showed everyone what it looks like
Today we’re releasing OpenUI 1.0 bringing it to every other agent
Same model: 24% on one harness, 59% on another.
That gap is the tooling around it. We only trust evals run on the stack we actually ship.
Model Benchmarks Are Worthless
How can a model score 24% and 59% on the same 113 coding tasks?
It happens when you vary the harness that wraps it.
I’ve increasingly ignored most of the recent model leaderboard results. Partly from exhaustion and partly because improvements in a ranking don’t show up in the actual work.
So I asked the 8090 Labs team to measure what the leaderboards skip: the model and the harness together, on real bug fixes and features from open-source projects.
8090 Labs ran 7 models through 5 harnesses.
Here are a few results I didn't expect:
- Anthropic's Claude Fable 5.1 scored 41% in Claude Code and 61% in OpenAI's Codex. It made 28 tool calls at the median in one and 66 in the other.
- The cheapest runs weren't the cheapest work. GLM in Codex cost $0.22 a run and $0.93 per success. In Pi it cost $0.25 a run and $0.43 per success.
- GLM passed a task on both attempts 51 times in one harness and 16 times in another.
- GPT-5.6 Sol and Grok 4.6 barely cared which harness they were in.
The bottom line is that if you're picking a model/harness combo off a model-only leaderboard, you're making a critical decision based on only half the information.
8090.ai/blog/same-model-diff…
No AI will replace the Tattoo Artist.
It can speed up the brief and help the client decide faster. That's where it stops.
The artist refines the final design before it touches skin. AI just gets you there faster.
Seven Adobe apps rebuilt as open source with AI.
I build with AI every day. The first 90% really is fast now. The last 10% is file formats, edge cases and years of boring fixes. That part is still the product.
AI is now being used to recreate Adobe’s software, a developer has released seven free, open-source apps as clean-room alternatives to Adobe’s main tools:
- PhotoCraft → Photoshop
- VectorCraft → Illustrator
- FilmCraft → Premiere Pro
- LightCraft → Lightroom
- EffectCraft → After Effects
- DesignCraft → InDesign
- PrintCraft → Acrobat Pro
Built in Rust with help from Claude Opus 5.5, they run on Windows, Linux, and macOS under MIT/Apache licenses without using Adobe’s source code.
Here we are 💀🖤
See it before it's permanent. Then hand your Tattoo Artist a clear brief, not a screenshot folder.
Watch how ↓
Most tattoo artists I know spend more hours on messages than on skin.
Quotes, chasing references, rescheduling. Nobody trained them for that part of the job. That is exactly what software should take off their plate.
Not the drawing.
Supabase Select was the real thing. Serious speakers, serious people in the room. People who know their stuff, people who do the work, people who have already built and are building at a ridiculous pace.
I met people from all over the world, and I'm sure I'll see plenty of them again and get to know them better.
Thank you @supabase, @kiwicopple and @AntWilson, not just for the event but for finding the time for a proper chat.
Ah, the #supasquad has brilliant guys and girls 😎
Two days setting up my whole workflow on my new @OpenAI Dot. Seriously impressive. For how I work, it beat even Grok Bot, which I've used since day one. The Spaces integration is brilliant.
Then I came back to europe (from the US), the magic vanished, and it was obvious why Europe is hopelessly behind.
@OpenAI, when is Dot coming to europe?
Taste came up in almost every good conversation I had in San Francisco.
Tattooing runs on it. A tool can rough out an idea in seconds. What goes on skin for life is the artist's call.
Inksquad is built around that: the tool sketches, the artist decides.
