How Phoenix works

Turn APOE4 risk into a plan you can trust.

One loop, four steps. It gets sharper every time you or anyone else runs it.

01Know

Four evidence sources, ranked into one shortlist for your genotype, with the uncertainty left visible.

02Act

One sequenced protocol, about 90 days, each step aimed at a single number on your panel.

03Proof

Bloods re-drawn against your own baseline. Your numbers, never an average.

04Together

Phoenix AI learns from 684+ carriers running experiments right now, and hands what it learns back to you.

684+ carriers · 150+ pods · 27 biomarkers tuned to APOE4
The Phoenix loopAlways running

Every result re-enters the loop: yours, and 684+ other carriers’ running right now.

Know

What may be worth trying, ranked.

Four evidence sources, compared against each other, so a promising idea rises to the top, and the uncertainty stays visible instead of getting rounded off.

  • PubMedPeer-reviewed evidence
  • Clinical trialsWhat's being tested now
  • APOE4 researchFindings specific to your genes
  • Other APOE4 carriersPatterns across the community
Phoenix AI compares the signalsWeights each source, flags where they disagree, and never hides a thin evidence base behind a confident sentence.
Your interventions shortlistRanked for your specific genotype and health profile.
Sources shown on every recommendationRe-ranked as new evidence landsHow Phoenix AI works
Act

One protocol, fully recorded.

Run the interventions your plan needs. Record the timing, dose, adherence, and target for each one, so every measurement has the context needed to identify what is helping.

Why record the whole stackThe point of the structure
Track every interventionMultiple changes are not the problem. Unrecorded changes are.
Every step has a targetA named biomarker, drawn from your own panel.
Your pod runs it with you2 to 4 carriers on the same step, checking in.
PhoenixYour protocol
Protocol layer 1 of 490-day measurement window
  1. SupplementsTargets your ranked biomarkers
    Now
  2. MovementBuilds toward your VO2 max
  3. SleepSteadies your HRV and recovery
  4. NutritionShapes your metabolic markers
Re-blood at day 90. Review every active intervention before the next decision.
How Phoenix AI works

Trained on evidence. Updated every week. Honest about what it doesn't know yet.

Phoenix AI is trained specifically on Phoenix member research, published APOE4 literature, and clinical trials — the same evidence sources and Know/Act/Proof/Together loop this page walks through.

Weekly refresh

The evidence base is updated weekly as research and trial information change. New does not automatically mean conclusive.

Uncertainty and limits

Phoenix shows caveats and what is still unknown. It helps members identify what may be worth testing and measure their own trends; it does not guarantee a result or replace clinician judgment.

Phoenix AI · Pattern detection

Five members. Five different stacks. One thing in common.

Each lowered LDL-C on a stack that looks nothing like the others. Phoenix AI reads across all five and surfaces the one intervention they share.

Member 1
APOE3/4 · 62 · F
14298mg/dL−31%
Vitamin D3
Creatine
Magnesium threonate
Ezetimibe
Lithium orotate
Member 2
APOE4/4 · 58 · M
163121mg/dL−26%
Glycine
Ezetimibe
Vitamin K2
Curcumin
Taurine
Member 3
APOE3/4 · 47 · F
12889mg/dL−30%
Lutein
Magnesium glycinate
Ezetimibe
NAC
Vitamin C
Member 4
APOE4/4 · 56 · F
171112mg/dL−35%
Ezetimibe
Vitamin D3
Vitamin B12
Folate
Melatonin
Member 5
APOE4/4 · 43 · M
151109mg/dL−28%
Creatine
Lithium orotate
Zinc
Ezetimibe
Citicoline
Pattern found
Ezetimibe
5 / 5
members share it
−30%
median LDL-C change
Illustrative members · intervention names from the Phoenix library
Phoenix AI · Members like you

You want to lower fasting insulin. Your twins already did.

Phoenix matches you to members who share your profile, then ranks what actually moved their number.

You
Female55APOE4/4HbA1c 5.7%Insulin 14 µIU/mL
12 members share your profile
9 lowered fasting insulin
What worked for them
ranked by twins who improved
Berberine
−31%9 of 12 twins
Zone-2 cardio · 3×/wk
−27%7 of 12 twins
Time-restricted eating
−24%6 of 12 twins
Alpha-lipoic acid
−18%5 of 12 twins
Magnesium glycinate
−9%4 of 12 twins
Illustrative · change = median fasting-insulin change across twins who logged the intervention and retested
Proof

Real members. Real numbers.

Not averages. Named carriers, their own markers, their own words, each one linked to the full interview so you can check it yourself.

Dr. Kevin Tran
Dr. Kevin TranAPOE4/4 · Doctor of Pharmacy · Founder
ApoB11570
VO2 max3951
HbA1c5.95.4
I built Phoenix because I needed it. Now we build it together.
Read Kevin’s full story
Robin P. Shwetzer
Robin P. ShwetzerAPOE3/4 · ReCODE coach · Florida
HbA1c5.75.4
Homocysteine13–147
I had to teach my own integrative medicine doctor about APOE4 — I knew more than the neurologists.
Watch Robin’s full story
Dr. John Yoder
Dr. John YoderAPOE3/4 · Retired physician · Missouri
p-tau0.09

Half the borderline threshold.

Even if they came up with a cure tomorrow, I would do everything I'm doing except the supplements. I love the working out.
Watch John’s full story
Dr. Gertrude Behan
Dr. Gertrude BehanAPOE4 · Retired physician · Queensland
Homocysteine down

On supplements targeted by her own genetic testing.

We're the wild variant. We're the tough humans.
Watch Gertrude’s full story
Deb Blum
Deb BlumAPOE4/4 · Health coach · Arizona
Weight down, muscle held

Post-menopause, glucose dialled in via CGM.

Do you want to just let your destiny be up to your genes, or do you want to have some agency?
Watch Deb’s full story
Donna Dorans
Donna DoransAPOE4/4 · 72 · 10 years since diagnosis
Runs a pod

Ten years in, still building her own protocol.

You can either sit back and do nothing or take action. That's still my attitude.
Watch Donna’s full story
0%

of members see improvements in APOE4 biomarkers within three months.

Read every member story
Start where you are

Your protocol starts with your data.

Join Phoenix. If you join, we'll show you where to start, based on your genotype, not the average adult's and not even the average carrier's.

We're building a protocol with the Phoenix app that's going to work for us, not just following someone else's protocol. You don't have to go to five different places.
Donna DoransDonna Dorans · APOE 4/4 · 72 · 10 years since diagnosis
Read Donna's full story →