Shaaf Kashan

Projects

Case studies, each broken down into context, constraints, decisions, trade-offs and outcome.

  • FRIDAY: Personal build: local-first AI assistant. Deciding, request by request, what is allowed to leave my Mac, and making sure the system actually follows that rule.
  • AI operating layer: Personal build: the setup I run my own work on. Letting AI work on almost 10,000 of my own notes every day without pasting them into chats, and without trusting a prompt to keep the rules.
  • TerraGrid: Final-year project: AI and blockchain property platform. Making a React frontend, an ML valuation model and a blockchain ledger behave like one product instead of three integrations side by side.
  • WealthSphere: Group project: financial planning platform. Putting budgeting, investment tracking and financial learning in one place that someone non-technical could actually use.
  • SWIFTBOT: First-year group project: robotics & software engineering in Java. Taking a robot task from requirements through design to a working Java implementation as a group.

Personal builds

  • Mihwar (iOS · on-device AI · local-first): A privacy-first daily-wellness iOS app with a 100% on-device AI companion. The language model runs on the phone, so a user's reflections never leave the device. Built solo in React Native; in development, with 622 passing tests.
  • ShaafAI (in-browser retrieval): The assistant on this site. A retrieval engine I wrote from scratch that answers from my own notes: no third-party AI, no API calls, all in the browser.