Signsbeat presents Yishii — the intelligence layer

The hardware exists.
The intelligence didn't.

Watches, rings, bands and CGMs watch beautifully. Yishii is the layer above them that understands — for employers, device makers, AI platforms and practitioners.

See how it works ↓ Talk to us

Scroll — the stack builds itself

How the layer works

Three layers, one intelligence.

Scroll through the layers — the signals already exist, Yishii is the layer that understands, and your interface wears the result.

Layer 3 · your interface
HR dashboard Device app AI assistant Practice
Layer 2 · Yishii intelligence
Reads · attributes · directs
Six callable modules. T-1 causal attribution. Plain language out.
Layers 0–1 · the signals
Apple Watch · Garmin · Oura · WHOOP · Fitbit · Samsung · CGM · food · labs · journal
Yishii AI · a live slice

The layer, in a user's hand.

Four saved questions, four different mornings. Scroll to move through them — or tap one. The member sees a conversation; your platform can read the structure.

y
Yishii
Today · 7:04am
Why am I so tired today?
Your nervous system didn't fully reset overnight — deep sleep ran 22% under your baseline, third short night this week. That's accumulation, not weakness.
The likely driver is yesterday: dinner at 9:40pm kept your glucose elevated past midnight. Tonight, eat before 8 and keep the evening quiet.
Ask Yishii…→
Yishii · daily read
Mild stress
61health score
−6 vs your baseline
What drove it
  • Deep sleep 48m · −22% vs baseline
  • Dinner 9:40pm → glucose high past 12am
  • Resting HR +4 bpm overnight
Do this
Eat before 8pm · keep tonight quiet
y
Yishii
Today · 7:04am
Big session planned. Can I push today?
Yes — today is a green window. HRV is back above your 30-day baseline and last night's deep sleep did its job.
Go before 11am, while recovery capacity is highest. Keep tomorrow light — two hard days back-to-back is where your pattern usually breaks.
Ask Yishii…→
Yishii · daily read
Recovery
84health score
+9 vs your baseline
What drove it
  • HRV 62 ms · +11% vs baseline
  • Deep sleep 1h 52m
  • Resting HR settled by 1:10am
Do this
Train hard, before 11am
y
Yishii
Today · 7:04am
Snapped my lunch — nasi lemak, extra sambal.
That's a high glycemic load for you — around 68. Expect a glucose peak near 2:15pm and an energy dip around 3, based on your own post-rice pattern.
A 12-minute walk in the next half hour flattens most of that curve. Next time: half the rice, double the cucumber — same plate, gentler afternoon.
Ask Yishii…→
Yishii · daily read
Post-meal
68glycemic load
peak est. 2:15pm
What drove it
  • GI est. 71 · glycemic load 68
  • Carbs 74 g · protein 18 g · fat 31 g
  • Your 30-day post-rice curve: +38 mg/dL
Do this
Walk 12 min before 2pm
y
Yishii
Today · 7:04am
My body feels off. Am I heading toward a flare?
Your pattern is drifting, not flaring. Resting heart rate has climbed for five days and HRV is trailing baseline — the same early shape as February, but milder.
This is the window where lifestyle still works. Protect sleep this week, keep training easy — and I've flagged the trend for your practitioner.
Ask Yishii…→
Yishii · daily read
Early pattern
57health score
5-day drift
What drove it
  • Resting HR +6 bpm across 5 days
  • HRV −14% vs baseline
  • Sleep fragmentation rising
Do this
Flagged to your practitioner · protect sleep
Four doors into the layer

Four ways in. One intelligence underneath.

01 · HR & corporate wellness

Wellness your people actually feel — under your brand.

A white-label dashboard your wellness team runs on day one. No native integration — Yishii reads the wearables employees already wear. Aggregate, opt-in, gently gamified.

Switch the brand — it's yours. →

Solstice Group Wellness
Powered by Yishii
1
People Ops
12 members
82+5
2
Product
34 members
78+4
3
Finance
19 members
74+2
4
Engineering
56 members
71+6
5
Sales
23 members
66−1
Opt-in · aggregated · anonymisedCompany streak · 12 days
02 · Hardware & API

You built the sensor. Yishii is the sense-making.

Smart rings, bands and watches gather extraordinary signals — then hand users a chart. Call Yishii's modules over API and your device starts answering the questions your users actually have.

Longevity Preventive health Women's health Autoimmunity
food_ai
One photo → glycemic index, load, macros, predicted response.
mental_fitness
Nervous-system state map with T-1 attribution. Fully passive.
metabolic_molecular
Inflammatory cell-signalling trend — direction before symptoms.
womens_health
Cycle-phase overlay on HRV, temperature, sleep, metabolics.
autoimmunity
Longitudinal early-pattern monitoring while lifestyle can act.
journal
Structured protocol log — practitioner-readable, private by default.
Powered by Yishii
Apple Health · Health Connect · Zepp OS · CGM
03 · AI & LLM platforms

Your model knows the literature. Yishii knows the user.

One API call returns structured, citable physiological context from the user's own signals — so your assistant stops answering health questions from the average.

Same question, same model. Flip the toggle. →

U
I keep waking at 4am and can't fall back asleep. What's going on?

Your last 14 nights show the 4am waking only on days you eat after 9pm — your glucose dips near 3:40am and your body treats it as an alarm. Eat by 8 tonight; for you this usually fades within three nights. This isn't insomnia.

  • Wearable — wake events 3:38–4:12am, 9 of 14 nights
  • Food log — dinner after 9pm on all 9 of those nights
— grounded in this body · one Yishii API call
04 · Clinicians, coaches & providers

The continuous picture between appointments.

Appointments last 20 minutes. Biology runs all week. Yishii gives the practitioner a continuous, plain-language read of a client's trajectory — a wellness-and-lifestyle lens, not another clinical system.

“The data my clients generate between sessions has always existed. Yishii is the first tool that tells me what it means.”

Living proof · autoimmunity

Cheryl Ban — pharmacist, health coach, SLE patient.

Early pattern shifts tracked across sleep, HRV and metabolic signals — lifestyle protocols adjusted before the body sent a louder signal. Published case; lifestyle protocol, not treatment.

Living proof · women's health

She was told IVF was her only option.

Through a monitored lifestyle programme, an early participant conceived naturally within the programme period. No claims about IVF outcomes — documented physiological optimisation.

Where the evidence stands

Built on a decade of data. Reviewed by researchers.

The evidence

The data already points one way.

A decade of continuous, N-of-1 biometric data sits behind the core algorithm — every user their own control. What's next is the paper, not the proof.

Under review

Validated with a research partner. Open to more.

Our approach has been validated with a university research group under an ethics-approved, observational design. We're open to more partners who want to work with the data.

Questions, answered straight

What people ask about Yishii.

What is Yishii?
Yishii is the intelligence layer above wearable health hardware. Watches, rings, bands and CGMs collect extraordinary signals; Yishii is the layer that reads across them, attributes cause to effect, and returns plain-language direction. It is built by Signsbeat, Singapore.
Who is Yishii for?
Four audiences: employers running corporate wellness, hardware and device makers, AI and LLM platforms, and clinicians, coaches and providers. Each reaches the same intelligence through a different surface — a white-label dashboard, an API, a model context layer, or a practitioner platform.
How does Yishii use wearable data?
Six callable intelligence modules read across every signal the hardware already collects, attribute cause to effect using the T-1 rule (today's state explained by yesterday's signals), and return direction in plain language. Every user is their own control.
What hardware and data sources does Yishii support?
Apple Watch, Garmin, Oura, WHOOP, Fitbit and Samsung devices, continuous glucose monitors, plus food, labs and journal entries. Device makers can integrate via Apple Health, Health Connect, Zepp OS and CGM data.
Is Yishii a medical device?
No. Yishii is a physiological pattern recognition and lifestyle intelligence platform. It is not a medical device and does not diagnose, treat or prescribe. All observational research is in progress.
How do partners integrate Yishii?
Hardware makers and AI platforms call Yishii's modules over API. Employers receive a white-label wellness dashboard. Clinicians and coaches use a practice platform that shows a continuous, plain-language read of a client's trajectory between appointments.
Start the conversation

Which door is yours?

Tell us who you are and what you're building on. We reply within two working days.

Or write to us directly — hello@yishii.app