Put the past
back together.
Reconstruct every antidepressant course from the record: drug, dose, start and end dates, anything added along the way.
From scattered notes to one timeline A clearer history. A considered next step.
Before the
next step,
check the last ones.
Hughlings turns a patient's medication history into an auditable review before the next escalation step — showing which antidepressant courses were really adequate, and why.
Research prototype / Tallinn, Estonia
From fragments
to a clearer picture.H—01
to a clearer picture.H—01
Every course. Every source.
Treatment-resistant depression · Escalation reviewScroll to discover
01 / The missing context
Three failed
medicines.
But what happened
in between?
The history that would show this is scattered across years of notes.
A patient is called treatment-resistant after several antidepressants fail. Sometimes the dose was too low, the course too short, or treatment was stopped early.
Reconstructing that history by hand takes time no one has.
A medicine on a list is only the beginning of the story.
01Dose
Was it high enough?
02Duration
Was it held long enough?
03Adherence
What supports it being taken?
04Exposure
What changed along the way?
02 / The approachFollow the thread
Input: the chart.
Output: clarity.
An auditable review.
One connected history.
Four deliberate steps.
A treatment history, reconstructed01 / 04
Course 01 Sertraline50 mg/day · 3 weeksDuration to review ↗
Course 02 Venlafaxine XR225 mg/day · 9 weeksExposure flag ↗
Course 03 Escitalopram20 mg/day · 8 weeksAdherence uncertain ↗
Week 4+ BupropionMetabolism to review
h.Review summary
Illustrative output
Illustrative output
Treatment history3 courses
AssessmentLinked to rules
EvidenceTraceable sources
Clinical decisionRemains with you
Illustrative interface · synthetic case
Put the past
Look beyond
the drug name.
Check each course. Was the dose high enough? Was it held long enough? What do pharmacy refills tell us — and what remains unknown?
Dose · Duration · AdherenceThe same dose.
A different story.
Flag drug interactions and metabolism that may have changed the real exposure, not just the prescribed dose.
Follow what changed along the wayEvery verdict.
A visible reason.
Stage the history with the Maudsley Staging Method and check it against the relevant reimbursement criteria. Every verdict links to its rule and source.
An auditable review, ready for a clinician03 / A closer lookA synthetic case
One chart.
One honest answer.
42 years old. Recurrent severe depression.
The chart says “three failed antidepressants”.
Look closer at each course.
DurationIllustrative review / 01
Three weeks.
A question left open.
The recorded course lasted three weeks. Duration needs to be assessed against the selected adequacy rule before this is counted as a failed adequate trial.
A short course does not tell the whole story.
Illustrative case. Not a real patient.
Hughlings doesn't say what to do next.
It shows what the history supports — and where it doesn't.
04 / Why now
The rules already ask
for this document.
Before the most resource-intensive escalation steps, reimbursement rules require a documented treatment history. Today psychiatrists assemble it by hand.
FIFinland / Kela 3062
A documented history.
A written statement.
Esketamine reimbursement requires a specialist statement and documented prior treatments, alongside severity and other criteria.
Read the Kela criteriaSESweden / Region Skåne
Every approach.
Every relevant detail.
Regional guidance asks for at least four previous treatment approaches. The referral template captures drug, dose, duration and concentration measurements.
Read the regional guidanceThe rule is the demand; esketamine is only the example.
Hughlings documents readiness. The clinical decision stays with the clinician.
05 / How we buildTrust is in the details
Informs clinical
management.
Does not drive it.
The clinician decides.
Hughlings does not prescribe, recommend or deliver treatment. The psychiatrist's decision is a separate, signed record.
Rules, not guesses.
Adequacy and exposure verdicts come from deterministic rules with named sources. A language model may explain a verdict; it never makes one.
Every output is auditable.
Each review records the exact input, the rule-set version and the time it was made. The same input and rule version produce the same result.
Designed for oversight.
We design for MDR and lifecycle oversight from day one, with a narrow intended use. Hughlings has not been clinically validated or CE-marked.
06 / Where we fit
Built on top,
not instead.
Medication safety is already national infrastructure. Evidence assistants answer clinical questions. Hughlings is building the treatment-history review layer between them.
We don't build interaction databases, pharmacogenomic reports, safety alerts or literature search. We build on them.
03The clinician's decisionIndependent. Considered. Signed.
02hughlings.History → Adequacy → Exposure → AuditThe review layer
01The existing foundationClinical records · Drug data · Evidence
Public psychiatric services preparing escalation decisions and reimbursement documentation.
Consistent, reviewable escalation histories across clinicians.
An escalation-review module for tools already inside the patient record.
07 / Where we areA work in progress, on purpose
Early.
And honest about it.
Hughlings is a research prototype.
The rules engine runs on a structured case, with every verdict sourced. It has not been clinically validated yet — that is what design partners are for.
Listen & review
Interviews with psychiatrists and clinical pharmacologists. Manual reviews of de-identified cases.
Build together
Work with 3–5 design partners.
First paid pilot
MDR classification with a fixed intended use.
Independent review
A blind benchmark against existing tools, scored by an independent assessor.
Founder
Yurii Sachenko
Software engineer. Three years shipping web products, three in product growth and positioning, and about five years studying psychopharmacology and pharmacokinetics.
An open seat at the table
We're looking for a clinical pharmacologist or psychiatrist to join as adviser or co-founder, ideally from Finland, Sweden or Estonia.
Start a conversation08 / Work with usFinland · Sweden · Estonia
Seeking 3–5 design partnersBring us five cases.
Bring us five cases.
We'll show you what
the history says.
You share a handful of de-identified cases. We review them together, by hand and with the prototype. You tell us where we're wrong.
Let's explore it together yura@sachenkolabs.techStart with a conversation. Please don't email patient information.