CG Common Ground | Peak Performance
What we did and what it producedActive

The work, decision by decision

What a member sees

One dashboard across every dimension that has data, with anything drifting out of range surfaced first. Eighteen-plus dashboard surfaces in four groups: core health (bloodwork, body composition, glucose, recovery and sleep), performance (training log, nutrition, supplements), deep analysis (genetics, sensitivities, AI-surfaced insights across datasets), and linked profiles (a member's own profile beside another profile they choose to link, compared side by side).

Capture takes lab PDFs, device exports, DNA files and plain free text; the system parses a note into a structured entry. Every chart traces back to a source document the member can open.

Genome comparisons

The vault has run five kinds of genome comparison, each described here as a method only:

  1. Two genomes compared against each other. Two people's genomes are compared side by side to see where they match and where they differ.
  2. A genome compared against a blood relative's. The same kind of comparison, run between two people who are related, to look at what they share by descent.
  3. Genetics and temperament. A genome is read against traits related to how a person relates to others.
  4. Genetics against supplements. A genome is read against a supplement protocol, so a choice can be checked against a genetic predisposition instead of guessed at.
  5. Genetics against sleep. A genome is read against sleep data, to see whether a genetic marker lines up with a sleep pattern.

The principles

Private by default. Tiered privacy on every record. Member-owned end to end, with full export built in and no data resale or aggregation. Single-tenant by design: each white-glove vault is its own deployment with its own database, so there is no pooled store and no honeypot.

Where it stands

A live private build, in daily use. One external white-glove build delivered, cloned end to end for a separate member. A public demo on fictional data, and a drafted commercial layer with no pricing wired yet. No entity exists yet; that decision sits at a later gate once a small run of external builds tests willingness to pay.

Questions answered

Member-Owned Health Data Vault

Walk me through one concrete integration or modeling decision where two sources represented the same health information differently. How did you decide what the platform should store and show?

It happens all the time. An Oura ring and an Apple Watch report the same measure differently, and Oura throws an outlier spike now and then. Lab panels do the same thing. So the Vault never treats one reading as the truth. It tracks trends and lets the outliers go. The trend lines usually move together even when the numbers don't. When they split, say Oura trending down while the Watch trends up over the same weeks, that is where a real problem is, and that is what we look at.

Answered 2026-09-29.

The strongest proof point is that you cloned the Vault end to end for an external member. What did that build demonstrate in practice, how long did it take, what did the member use it for, and what changed because they had the unified view?

It showed the Vault travels. I've cloned it for about six members now. The build itself takes about an hour. The slow part is the member: gathering their blood work, genetic tests and supplement list usually takes about a week while they dig for it. Most people don't have the depth of data I do, since I run DEXA scans and more blood and genetic testing, so a clone usually starts with about 30 percent of what my Vault holds. What each member does with it is theirs and stays private. The point is the structure works without me rebuilding it.

Answered 2026-09-29.

What it produced

A live, private build in daily use, one external white-glove clone delivered end to end, and a public demo running on entirely fictional data. The venture framing, pricing and roadmap remain a working draft.

A slice of the project list

A few related projects.