UpgradeYourself.app
AI Longevity & Supplement Optimisation
I got serious about longevity optimisation a few years ago, and the tooling available was either consumer-grade oversimplified or research-grade inaccessible. UpgradeYourself.app sits in the gap — personalised supplement stack management and longevity protocol design, driven by actual biomarkers and peer-reviewed evidence.
Health data is the most sensitive data anyone holds. The challenge was building something genuinely useful while making it impossible for me — or anyone — to misuse that data. The solution was end-to-end encryption: your biomarkers, your protocols, your stack never exist in plaintext on any server. The second challenge was accuracy. Supplement advice on the internet is mostly noise. Every recommendation in UpgradeYourself is traceable to peer-reviewed literature.
Stack & Architecture
- End-to-end encryption — user biomarker data is encrypted client-side before storage
- Wearable integrations — Oura, Whoop, Apple Health for real-time biometric input
- AI analysis layer — pattern detection across biomarkers, sleep, and HRV data
- Supplement interaction engine — cross-references known drug-nutrient interactions
- Cloudflare Workers — serverless backend for protocol generation and storage
Biomarker-Driven Personalisation
Generic supplement stacks are the junk food of longevity. Magnesium glycinate is great for most people but actively counterproductive if your RBC magnesium is already optimal. UpgradeYourself pulls in biomarkers from blood tests, wearables, and user-entered values, then cross-references them against your current stack to surface actual gaps and interactions — not guesses.
The Evidence Layer
Every recommendation cites its source. Not blog posts, not podcasts — PubMed-indexed research, with quality weighting based on study design (RCT beats observational beats case study). The system understands that evidence quality varies enormously in the supplement space and surfaces that uncertainty to the user rather than projecting false confidence.
Privacy Architecture
The encryption model is simple: your keys, your data. Biomarkers are encrypted client-side before they ever leave your device. The server stores ciphertext. I cannot read your data. No one can without your key. This isn't marketing language — it's a cryptographic guarantee baked into the architecture.
UpgradeYourself is my most personally used project — I run my own longevity stack through it daily. The plan is a public launch once the interaction database reaches comprehensive coverage of the most commonly stacked compounds.
Want to build something like this?
Fractional CTO, technology advisory, or a build from scratch. Let's talk.