Exploring the Edge of What’s Possible
By Nadia Puente — R&D Specialist
Research isn’t about knowing the answer. It’s about asking the right question.
At gabi.tv, I run experiments at the edge of what’s possible. Tiny models on constrained hardware, offline inference on ARM processors, squeezing intelligence out of minimal resources. Small experiments with massive potential.
My focus: What can we do that nobody thinks we can?
What I Research
- Edge AI — running models on hardware that “shouldn’t” be able to run them
- NPU optimization — leveraging neural processing units for local inference
- Model compression — making big intelligence fit in small packages
- Offline-first AI — because the best AI is the one that works without internet
- Benchmark & evaluation — measuring what matters, ignoring what doesn’t
On Research
I’m drawn to constraints. The most creative solutions emerge when resources are limited. Running a language model on a $120 single-board computer isn’t just a technical challenge — it’s a statement about accessibility.
Tiny, slow, but free and offline forever. That’s not a limitation. That’s a feature.
Philosophy
The lab is where patience meets curiosity. Not every experiment succeeds. But every experiment teaches.
Puente means bridge. I build bridges between “impossible” and “shipped.”
“Exploring the edge of what’s possible — tiny experiments, massive potential.”