Now piloting with universities and employers in Delhi NCR — read about the pilot programme
Clinical decision support · Women's health

From symptoms to answers.

Women’s health, simplified.

Seven minutes to explain seven years. The intelligence layer between experiencing a symptom and receiving appropriate care. Valkyria turns a woman's lived history into structured clinical evidence her doctor can act on — and tells her, clearly, when it is time to go.

See how it works
Clinician-in-the-loop Hormone-aware Multilingual & low-bandwidth ISO 13485-aligned EMR-light

Built with practising gynaecologists · India-first

Pre-consult summary Female, 47 · 214 days logged · prepared 12 Sep REVIEW ADVISED BLEEDING & PAIN — LAST 7 CYCLES Mar Sep INSTRUMENT SCORES Menopause Rating Scale24 / 44 PBAC (menstrual blood loss)118 — elevated PHQ-97 / 27 INDICATED FOR REVIEW Full blood count with ferritin Thyroid function Cervical screening — none on record
The problem

Women don't delay care because they don't want care.

They delay because nobody has told them when a symptom crosses the line from ordinary to worth investigating. The condition progresses in that gap.

01

She doesn't know when it matters

Heavy bleeding, fatigue, pain. Normal, or disease? No one has ever given her a threshold.

02

She doesn't know where to go

GP, gynaecologist, endocrinologist, a lab test? The system assumes she can route herself.

03

She doesn't know if it's worth it

Told once that it's stress, she recalibrates downward and stops going back.

These conditions are diagnosed by pattern, over months. The system gives her seven minutes and asks her to remember.

Use cases

She never pays.

No subscription, no paywall, no premium tier for the woman using it. Institutions pay instead, because they are the ones already carrying the cost of late diagnosis — in absence, in attrition, in avoidable escalation.

  • Employers. Annual checks for staff over 40 are now statutory, and most go unused. We lift participation and turn a generic panel into an indicated one. Read more →
  • Universities. Campus health sees irregular cycles, anaemia and distress constantly, with no longitudinal record. Read more →
  • Hospitals & clinics. A gynaecologist seeing fifty patients a day can't build a history in the room. We build it before the patient arrives. Read more →
Clinical scope

Narrow, deliberately.

Validation before breadth. We start where diagnostic criteria are established, tests are inexpensive, and delay does the most damage.

Perimenopause & menopausal transition Iron deficiency & heavy menstrual bleeding Thyroid dysfunction PCOS & metabolic risk Cervical screening follow-through Endometriosis — in development
A woman and her gynaecologist reviewing her symptom history together during a consultation
Get involved

Early access, and a team to build.

We're onboarding a first cohort of institutional partners and the clinicians who'll define what Valkyria flags.

Clinical roles
Product · The layer

Clinical decision support

The core of Valkyria. It watches a symptom history accumulate and identifies the moment it crosses a threshold that published clinical criteria say warrants investigation — and which investigation.

How it works

  • Score, don't guess. Reported symptoms map onto validated instruments rather than bespoke heuristics.
  • Threshold, not sentiment. Escalation fires on defined criteria, and every flag carries the guideline it came from.
  • Name the next step. Not "see a doctor" — which doctor, and which test is indicated.
  • Clinician in the loop. The system organises and surfaces. A human decides.

Instruments we build on

  • Menopause Rating Scale — menopausal symptom burden
  • PBAC — semi-quantitative menstrual blood loss
  • Rotterdam criteria elements — PCOS assessment
  • PHQ-9 and GAD-7 — mood and anxiety, as routing inputs
What it is not. Valkyria does not diagnose, does not prescribe, and does not replace examination. It decides when a clinician should look — and makes that look faster.
Regulatory posture

Built to be certified, not to slip through.

Software that interprets physiological data for clinical decision-making is a medical device, and we're building as though ours already is: quality management and risk processes aligned to ISO 13485 and ISO 14971 from the first commit, with a clinical evaluation plan written before the code.

We market a clinician-in-the-loop configuration while the Indian validation evidence accumulates, and file once it does. Retrofitting compliance is how health software companies lose two years.

Product · Capture

Adaptive symptom logging

Most tracking apps ask everyone the same questions forever. Ours changes shape based on what she has already reported — light by default, and considerably more thorough the moment something looks clinically relevant.

Light by default

A few taps a day. No questionnaire, no streaks, no guilt mechanics.

Deeper when it matters

Three heavy cycles in a row and it starts asking about clot size, flooding and fatigue — the details that separate a nuisance from iron deficiency.

Built for Indian reality

Multilingual, low-bandwidth, offline-tolerant, usable on a shared or entry-level phone.

Architecture

Ten reusable clinical domains.

Logging is not organised around conditions. It is organised around body systems — so the same structured record supports decision support across every condition we add, without rebuilding capture each time.

01

Pain

Site, severity 0–10, quality, triggers, relation to menses and activity, response to analgesia.

02

Fatigue & energy

Severity, post-exertional worsening, sleep quality, daytime sleepiness, exercise tolerance.

03

Mood & anxiety

PHQ-9 and GAD-7 at defined intervals, day-to-day mood rating, and safety escalation.

04

Sleep

Duration, latency, night waking, non-restorative sleep, circadian shift.

05

Gastrointestinal

Appetite, nausea, Bristol stool pattern, bleeding, relation of pain to meals and defecation.

06

Neurological

Headache pattern, dizziness, weakness, numbness, vision and speech change.

07

Cardiorespiratory

Breathlessness, orthopnoea, chest pain, palpitations, oedema, weight change.

08

Genitourinary

Frequency, urgency, dysuria, incontinence, pelvic pain, full menstrual characteristics.

09

Dermatological

Rash, photosensitivity, hair loss, acne, hirsutism.

10

Functional impact

Work and study attendance, activities of daily living, exercise tolerance, social functioning.

Why domains rather than conditions. A condition-shaped tracker only ever finds what it was built to look for. Domain-shaped capture means a woman logging for perimenopause is simultaneously generating the record that would surface thyroid disease, anaemia or a cardiac presentation — and it means adding a condition is a rules change, not a new product.
Safety escalation. Where a response in the mood domain indicates risk of harm, the system does not file it into a case summary for later review. It responds immediately, surfaces crisis support, and routes to a human. A flag that waits for the next appointment is not a safeguard.
Product · Escalation

Red flags & routing

Knowing something is wrong is useless without knowing where to take it. Valkyria tiers urgency and routes to the right kind of clinician — then keeps checking that the referral actually closed.

  • Urgent. Presentations warranting same-day assessment. Stated plainly, without hedging or alarm.
  • Investigate. A pattern meeting criteria for workup within weeks, with the indicated tests named.
  • Monitor. Worth watching, not worth a consultation yet. Explicitly told so, which prevents both panic and neglect.
  • Baseline. Nothing crossing threshold. Also worth saying out loud.

Routing and closure

Routing is specialty-aware: heavy bleeding with fatigue goes somewhere different from cyclical mood change. Where an institution has its own referral map, we use theirs.

Then the part everyone skips — we follow up. In Indian public and private screening alike, the referral that never completes is the commonest failure. A flag that nobody acts on is not a feature.

Product · Handoff

Consultation summary

One page, in the clinician's hands before the patient sits down. The design constraint is absolute: if it adds thirty seconds to a consultation it will never be used, so it has to save five minutes.

What's on it

  • Dated symptom timeline, not a narrative
  • Instrument scores with elevated values marked
  • Red flags, with the criteria that triggered them
  • Prior results and what's missing from the record
  • Indicated tests, with guideline reference
  • Year-on-year comparison once history exists

How it reaches them

  • EMR-light. Works alongside paper, WhatsApp, and whatever the practice already runs. No migration required.
  • Print, PDF or screen. Formatted to be scanned in twenty seconds, not read in three minutes.
  • Configurable. Hospital groups get the layout and fields their clinicians prefer.
Product · Capture

Cycle & hormonal tracking

Cycle tracking is table stakes and we treat it that way — it isn't the product, it's the baseline everything else is scored against. Without knowing her normal, no threshold means anything.

Her baseline

Cycle length, variability, flow, and symptom timing across the month — the reference against which change becomes detectable.

Change detection

Shortening cycles, rising variability and new symptom clusters are among the earliest signals of the menopausal transition, and are routinely missed.

Not a fertility product

We don't predict ovulation or sell conception windows. Different problem, well served elsewhere.

Product · The agent

Meet Saga.

A saga is a record of events kept over time. Saga is one clinical assistant with two surfaces — the part a woman talks to, and the part her doctor works in. Both read the same record.

For patients

The part she talks to

Saga asks about symptoms in plain language, in her language, and builds the longitudinal record as she goes. It watches for the patterns that matter and tells her clearly — without alarm and without dismissal — when something is worth a doctor’s time.

Women describe symptoms in stories, not form fields. Saga takes the story and produces the structure, which is exactly the translation the health system currently asks the patient to perform alone, under time pressure.

For clinicians

The part they work in

Saga drafts the case file before the consultation: history, instrument scores, flags, prior results, indicated tests. The clinician opens it by the patient’s Valkyria ID and revises it through chat rather than forms.

“Add suspected adenomyosis. Ordered pelvic ultrasound. Review in six weeks.” Saga writes it into the file, sets the follow-up, and the patient’s record reflects it before she has left the room.

Case files

One record, one ID, both sides.

Every woman has a single Valkyria ID. Her history, her flags and every clinician note live against it — so the next consultation, at a different clinic in a different year, does not start from zero.

  • Drafted, not dictated. Saga prepares the case file from the logged record. The clinician owns it and can rewrite any part of it.
  • Revised by chat. No forms, no template fields. A clinician seeing fifty patients a day will type a sentence; they will not fill a screen.
  • Versioned and attributable. Every revision is tracked, timestamped and attributed. What the software proposed and what the clinician decided stay distinguishable — which matters clinically and legally.
  • Consent-bound and portable. The record belongs to the patient and moves with her, under her consent, in line with Indian data protection obligations.
  • Corrections improve the system. When a clinician overrides a flag, that disagreement is the most valuable signal we collect. It is how thresholds get calibrated to Indian practice rather than imported wholesale.
What Saga never does. It does not diagnose, prescribe, or overrule a clinician. It decides when a human should look, prepares what that human needs, and records what they decided.
In development

Wearables as passive signal

The most reliable data is the data she doesn't have to enter. Where a woman already wears a device, we want to read the signals that correlate with what she's reporting — and notice when the two disagree.

What we'd use

Resting heart rate, heart rate variability, skin temperature, and sleep architecture — all of which shift across the cycle and the menopausal transition.

Devices

Apple Watch and Oura first, via their health platforms. Others where an accessible, consented data pathway exists.

Honest caveat

Consumer wearable data is noisy and not diagnostic. It earns a place in the clinical picture only once we can show it improves detection — not before.

Status. Not in the current product. Scheduled after the core decision-support layer has Indian validation data, because a feature that sounds impressive and changes no outcome is a distraction.
Clinical scope

Where we start, and why.

A decision-support tool is only as good as its evidence base, so breadth is earned. We begin with conditions where diagnostic criteria are established, the confirming test is inexpensive, and the cost of delay is high.

  • Perimenopause and the menopausal transition. Average age at menopause in India is materially earlier than in the West, placing it squarely inside working life — with validated symptom instruments and almost no clinical pathway.
  • Iron deficiency and heavy menstrual bleeding. Anaemia among Indian women is extraordinarily common and routinely treated without ever investigating the cause. Heavy bleeding to iron deficiency to dismissed fatigue is the most under-examined chain in Indian women's health.
  • Thyroid dysfunction. Inexpensive to test and it mimics almost everything else we track — fatigue, weight change, cycle irregularity, low mood.
  • PCOS and metabolic risk. High prevalence, high search intent, and a long-term metabolic trajectory that makes early identification genuinely consequential.
  • Cervical screening follow-through. India's screening coverage remains a fraction of eligible women, and a positive result that never reaches confirmation is a preventable death.
  • Endometriosis — in development. The starkest diagnostic delay of all, but confirmation is surgical and Indian primary-care awareness is limited. We're building toward it rather than claiming it.
  • Perimenopausal mood disorders — in development. Frequently treated as primary depression when the hormonal trigger is never identified. Narrow, validated, and squarely within scope.
Use case · Employers

Make the mandate actually work.

You are already required to fund annual health checks for employees over 40. Industry utilisation of health-check benefits sits at roughly a third — and for women in that cohort, the standard panel finds almost nothing specific to them.

What we deploy

  • Structured symptom intake before the blood draw
  • Indicated panel selection instead of a one-size default
  • Results interpreted against her reported symptoms
  • Booked follow-up with a clinical brief attached
  • Year-on-year comparison as the record builds

What you get

  • Utilisation you can report. Participation among the cohort that currently ignores the benefit.
  • Compliance evidence. Documented delivery, not just a vendor invoice.
  • Population view. Anonymised, aggregate — what your workforce is actually presenting with.
  • Retention signal. The 40–55 female cohort is where organisations quietly lose senior women.
How we sell. Usually through the benefits platform you already use, as a women's health module rather than another vendor to procure. If you'd rather contract directly, we do that too.
Use case · Universities

Catch it where it first presents.

Campus health services see irregular cycles, anaemia, pain and distress constantly — with no longitudinal record, no triage layer, and a student population that mostly doesn't present at all until something is acute.

For students

A private first step. Log symptoms, understand whether they're within normal range, and get told plainly when to see someone — with the record travelling with her when she does.

For the health centre

Students arrive with a structured history instead of a vague account, which shortens consultations and improves referral quality.

For the institution

Aggregate, anonymised visibility into what the student population is presenting with — and evidence for where to direct welfare resource.

Why universities matter to us. Honestly: this is where our clinical evidence comes from. Campus cohorts give us consented longitudinal data faster than any other setting, so university partners get the product free and permanently.
Use case · Hospitals & clinics

Configurable decision support, built around your bottleneck.

Every practice has a different problem. A fertility group cares about time-to-workup. A multi-speciality chain cares about referral leakage. A screening programme cares about loss to follow-up. The engine is the same; the thresholds, pathways and outputs are yours.

  • Custom pathways. Your protocols, your escalation thresholds, your referral map.
  • Pre-consult briefs. Formatted for how your clinicians actually read, not how software wants to print.
  • Follow-up closure. Track referred patients until the loop closes, with the drop-offs visible.
  • Population view. What your cohort is presenting with, before it reaches OPD.
  • EMR-light. Works alongside paper and whatever system you already run.

Book a clinical demo

Thirty minutes, with a clinician in the room. We'll walk through a real pre-consult summary for a case type you choose, and be direct about what the product does not yet do.

Commercial note. Where diagnostics are involved, any revenue arrangement sits between Valkyria and the institution or laboratory — never with a referring clinician. We won't structure anything that touches a doctor's referral incentive.
Pilot programme · Open

Six months, free, with a say in what we build.

We're selecting a small first cohort of employers, universities and clinics in Delhi NCR. You get the product at no cost; we get the clinical evidence and the honest feedback. Both sides need this to be real rather than ceremonial.

01

Scope, together

Two weeks. We map your current pathway, define what success would look like, and agree what we'd measure.

02

Run it

Six months live with your population, your clinicians, and a named point of contact who is one of the founders.

03

Read the result

What was found that would otherwise have been missed, what the clinicians actually used, and what didn't work. Published to you in full.

If you're an individual, not an institution. Join the waitlist anyway. We don't currently sell to individuals, and the product is free for the women who use it — but knowing where demand is coming from is how we decide which employers, campuses and clinics to approach next.
What we ask in return. Access to clinicians for design feedback, permission to use anonymised aggregate outcomes in our clinical evidence base, and a reference conversation if it goes well. Nothing else.
About

Why we're building this.

Ask almost any woman about the last time a doctor told her the pain was normal and she'll have an answer ready. That isn't a failure of individual clinicians. It's structural.

Women's conditions are diagnosed by pattern over months. The consultation is seven minutes. Nothing in the system holds the pattern, so the burden of carrying it falls on the patient — from memory, under pressure, to someone with no prior record of her. The people least equipped to win that argument are the ones who go undiagnosed.

Valkyria exists to hold that record, and to be the thing that says clearly: this is worth a doctor's time. We don't charge her for it.

A woman and her gynaecologist reviewing her symptom history together during a consultation
Founders

Alisha Chadha & Priyanshi Grover

Building Valkyria out of Delhi. We are pre-launch, running clinical validation interviews and preparing our first institutional pilots — and we answer our own email.

Contact

Clinically grounded

Validated instruments, guideline-referenced thresholds, and practising gynaecologists defining what we flag.

India-first

Multilingual, low-bandwidth, EMR-light, and built for a health system that runs on paper and WhatsApp.

Free for her, always

Institutions pay. A woman should never have to buy the right to be taken seriously.

Careers

We're looking for doctors first.

The hardest part of this product isn't the software. It's deciding where the thresholds sit — and we'd rather build those with clinicians who want to define them than inherit someone else's.

Clinical

Gynaecologists & clinical advisors

Define escalation criteria, review pre-consult brief design, and co-author the validation work. Named advisory roles with equity, not an unpaid logo on a slide. Endocrinology and general medicine backgrounds equally welcome.

Write to us →
Engineering & design

Founding engineers

Building a regulated clinical system for low-bandwidth, multilingual, shared-device reality. If you've shipped health software under a quality management system, we especially want to hear from you.

Write to us →
Where we are. Pre-launch, Delhi, running validation interviews and preparing our first institutional pilots. Early enough that what you join shapes what it becomes.
Waitlist

Be in the first cohort.

Tell us where you're writing from and we'll come back to you with the version of Valkyria that's relevant — institutional pilot, clinical role, or early product access.

Takes about thirty seconds — name, phone and email.

Institutional pilots

Employers, universities and hospital groups. Six months free, your own population dashboard, and a say in the roadmap.

Enquire about a pilot →

Doctors & clinical advisors

Help define the escalation thresholds rather than inherit them. Named advisory roles and co-authorship on validation work.

Write to us about a clinical role →
Product · Personalisation

Hormone-aware care

The same symptom means different things depending on where a woman is in her cycle, whether she is in the menopausal transition, and whether she is on hormone therapy. Almost no digital tool accounts for this, which is why generic symptom tracking produces noise instead of signal.

Cycle phase

Heavy bleeding on day three and spotting on day twenty-one are different clinical events. Fatigue confined to the luteal phase is not the same finding as fatigue that never lifts. Every entry is positioned against her own cycle, not a calendar.

Life stage

A hot flush at 47 with lengthening, variable cycles is a different question from the same symptom at 32. Stage determines which instrument applies — and which thresholds are clinically meaningful.

Hormone therapy

On menopausal hormone therapy, the expected symptom trajectory changes. A flag that ignores her prescription is a false alarm, and false alarms are how clinicians learn to ignore a tool.

What personalisation means here. Not a population average applied to her. Her own baseline, established over months, with deviation from it as the signal. Two women can report identical symptoms and only one of them is changing — that difference is the entire point.
Platform

The clinical data foundation

Decision support is only as good as the data underneath it. Most health data in India is transactional — a visit, a bill, a report — captured at the facility and unstructured at the point that matters. Ours is structured at capture, scored on entry, and continuous.

What makes it usable

  • Structured on entry. Ten clinical domains, fixed fields, validated instruments — not free text to be parsed later.
  • Granular and longitudinal. Daily resolution over months, positioned against each woman’s own baseline and cycle.
  • Clinically labelled. When a clinician confirms or overrides a flag, that judgement is recorded against the record. Labelled data is the scarce input in clinical AI, and it arrives as a by-product of use.
  • Consent-bound. Patient-owned, purpose-limited, and built to India’s data protection obligations from the outset.

Why it compounds

Every month of use improves threshold calibration for Indian practice, which improves flag precision, which earns more clinician trust, which produces more labelled overrides. That loop is the asset. It cannot be bought, licensed or shortcut — only accumulated.

What we do not do. We do not sell patient data, and we do not treat it as a product. It exists to make the decision support work, to support validation, and — with explicit consent and ethics approval — to contribute to research on conditions that have been studied far too little.
Product · Care model

Remote & hybrid care

Care for these conditions is mostly not delivered in a consulting room. It happens in the months between visits, and increasingly through sample collection at home. Valkyria is built for that reality rather than around the appointment.

Between-visit management

After a consultation, the plan usually disappears. We keep the thread — scheduled check-ins, symptom response to treatment, and a flag if things move the wrong way before the review date.

Home sample collection

Home phlebotomy is already routine through Indian diagnostic chains. We identify the indicated panel and hand off to the collection partner; results return into the record and are read against her symptoms.

Telehealth, integrated not owned

We route into the consultation platforms clinicians and employers already use, carrying the pre-consult brief with her. Building another doctor marketplace would help nobody.

Home diagnostics: honestly, not yet. There is currently no validated at-home diagnostic test for endometriosis, PCOS or breast cancer. A research group at Aberystwyth University is developing lateral-flow urine tests for all three and hopes to have a breast cancer prototype within about a year; a University of Cincinnati team is working toward a first at-home endometriosis test. Systematic reviews still find no reliable non-invasive diagnostic tool for endometriosis. We are tracking this closely and will say so plainly when that changes.
Why it matters to us. A biomarker result without symptom context is just a number. If and when validated home tests arrive, the layer that reads them against months of longitudinal history is where they become clinically useful — which is the layer we are building now.
Founders' contact

Contact the founders.

We are a two-person company and we answer our own phones. If you would rather skip the form, write or call either of us directly.

Co-founder

Priyanshi Grover

Co-founder

Alisha Chadha

Company enquiries. biocanvasprivatelimited@gmail.com · Delhi NCR, India. Waitlist and demo requests submitted through this site land in the same inbox.