AI & ML interests
https://github.com/EdgePhoneAI01 https://github.com/NonExecAI01 https://github.com/GreenH2sAI01
Recent Activity
et.ai
Coder · AI/ML neural networks · Business developer · Patent holder
About
I build AI products that have to work in the real world: on a phone, in a board pack, and across green energy systems.
My public work sits at the intersection of on-device inference, governance intelligence, and green fuels. I hold patents, ship product, and keep a working lab of the models, agents, and edge stacks I want inside those products.
- Building EdgePhone.ai — phone-first edge AI that brings fast, private, and reliable machine intelligence onto devices; apps feel instantaneous, user data stays local
- Building non-exec.ai — an AI tool to support Non-Executive Directors
- Building greenh2s.ai — AI for green energy and green fuels
How I work: prototype on the edge, measure against the market, then productise. Code is the lab. The companies are the shipping lane.
What I am shipping
EdgePhone.ai
Phone-first edge AI
Fast, private, and reliable machine intelligence on the device. Apps feel instantaneous. User data stays local.
Website · GitHub · LinkedIn · X
non-exec.ai
AI for Non-Executive Directors
An AI tool to support Non-Executive Directors with board-level insight, without handing the job to the executive pack.
Website · GitHub · LinkedIn · X
greenh2s.ai
AI for green energy and green fuels
Applied AI across green energy and green fuels — systems that have to clear a project, not only a slide.
Portfolio org: AIPortfolio · GreenH2s · NonExec · EdgePhone
How I think about the stack
devices ──► on-device / 1-bit / edge inference
│
agents ──► specialised crews, evals, and copilots
│
boards ──► governance intelligence, not dashboards for their own sake
│
energy ──► AI for green energy and green fuels that has to clear a project, not a slide
Working set: Python · TypeScript · on-device ML (BitNet, Google AI Edge, LocalAI) · agent systems · Android / Kotlin · product + patent development
In the lab
Public GitHub is a working notebook, not a trophy case. I fork, break, and reassemble the stacks I intend to productise.
| Area | What I am studying | Repo |
|---|---|---|
| On-device GenAI | Local gallery of edge ML / GenAI use cases | google-ai-edge-gallery |
| 1-bit LLMs | Official BitNet inference for 1-bit models | BitNet |
| Local runtime | Run LLMs, vision, and voice without a GPU tax | LocalAI |
| Agents | Specialist agent crews with process, not just prompts | agency-agents |
| Reasoning | ARC-AGI toolkit and train/test corpus | ARC-AGI |
| Market signal | Ranking checks across the three product domains | rankingcheckaimodels |
Connect
Source profile: github.com/etai-netizen · etai-netizen/README.md