Builder profile · privacy-first systems · real-world automation

I’m trying to change the world now.

Not later. Not as a vague ambition. Now. I build practical systems around automation, privacy, local AI, research intelligence, and personal infrastructure. If this resonates with you, I’d love to discuss it.

What I build, what I believe, and where I’m going.

I once heard Jensen Huang say we should do something the world has never done before. That idea stayed with me.

So I started building things I actually needed: an automated badminton booking system, a research subscription workflow for academics, privacy-first local assistant infrastructure, and tools aimed at real-world usefulness rather than demos for their own sake.

I’m now at a point where I could keep going entirely on my own, or join a company whose soul, technical ambition, and long-term view genuinely match mine. For now, I want to talk to people who are serious about building the future properly.

Projects from actual needs

Built

Badminton Booking System

An automatic booking workflow for badminton courts, built for me and my friends. No business model. Just useful software that removes friction from life.

  • 5 active users
  • Practical automation over polished theory
  • Useful because it solves a real repeated problem
Live direction

Research Subscription

A research monitoring workflow designed for academics and research leads: conferences, journal special issues, workshops, submission opportunities, and high-impact papers.

  • Weekly or biweekly briefing model
  • Built for me and my academic supervisors
  • Early form of a broader research intelligence product
Ongoing

Fieldcraft / “LeetCode for FDE”

A developing framework around field engineering discipline, practical judgment, and problem-solving under real constraints.

  • Still in progress
  • Focused on capability, not performance theater
  • Built around long-term practical skill
MVP

Local Personal Assistant Architecture

A privacy-first assistant system where sensitive data stays local instead of going to the cloud. The phone acts as a sensor and interface. The brain is hosted locally and connected via VPN.

  • VPN finished
  • First Android phone-to-brain MVP finished
  • Next step: closed-loop integration

Ideas I think are worth pursuing

AI token infrastructure

I’m interested in the market around model usage itself: token distribution, routing, pricing, and how different latency classes should map to different compute and geography.

  • Token distributor models
  • Latency-aware token routing
  • Cheaper long-running work vs ultra-low-latency local work

Enterprise local AI

Many organizations in the UK, Europe, and the US care deeply about privacy, compliance, and data protection. That creates room for local model deployment and controlled upgrade pathways.

  • Compliance-first model implementation
  • Local enterprise inference
  • Safer upgrade and fine-tuning paths with less leakage risk

Personal infrastructure

I believe extremely sensitive personal data should be controlled by the individual, not defaulted to the cloud. That includes assistants, cameras, notes, context, and memory.

  • Private VPN for individuals
  • Local-first personal data control
  • Personal large models and true personal assistants
  • Phone as a server-side participant in daily life systems

Useful systems, not empty theatre

Build things that solve real friction.
Treat privacy as architecture, not branding.
Prefer local control for sensitive data.
Think in systems: latency, compliance, cost, trust.
Work with people who want to build seriously.

I’m trying to change the world now.

If you are building something ambitious, useful, privacy-conscious, or genuinely future-facing, I’d be glad to discuss it.