A LONGITUDINAL LAB VISUALIZATION PLATFORM · FHIR-NATIVE · LOINC-CODED

Visualize faster. Chart by voice. AI-modeled trajectories tomorrow.

TrendHUD makes care faster, more accurate, and more convenient — without adding work for the physician.

The problem

Lab data lives in lists.
Patients live in trends.

Reviewing a year of CBCs in Epic means dozens of clicks across dozens of result panels. TrendHUD ingests every LOINC-coded lab from your EHR export and renders the whole trajectory on one screen — correlations without the click-storm.

Correlations without the click-storm
How it works

A journey from raw FHIR export to clinical insight.

One canvas. Three waypoints. Drop the file, fly forward.

1 Ingest

Pulls every LOINC-coded observation from FHIR exports.

Epic, NextGen, InteliChart — drop the file, we parse it.

2 Trend

Every metric, every time point, on one canvas.

Normal-range bands rendered as translucent green planes so abnormal values pop.

3 Interpret

An AI agent reads the trends with you.

Surfaces the data of clinical interest — and proposes the next step, grounded in the patient's own trajectory.

Live demo

This is TrendHUD.

An 18-month IDH1-mutated AML chart. Toggle metrics in the sidebar. Hover the timeline to crosshair every panel at once.

Hit the blue brain to see the AI agent.

Real product UI. Mock patient (de-identified). Every line, every band, every cycle marker is exactly what a TrendHUD physician sees.

Interpret · AI analytics

Trained on the way oncologists actually read a chart.

Click Interpret and the agent reads your patient's longitudinal record the way a colleague would — pointing to the trends worth your attention, citing the dates that support them, and proposing the next move grounded in your own protocol.

01

Surfaces what matters first

Ranks abnormal trends by clinical significance, not row order. The dip in your ANC before the next cycle gets surfaced — not the in-range RDW.

02

Cites the chart, not the cloud

Every observation points back to the data points and source dates that triggered it. Click a citation, jump to the timepoint.

03

Speaks your protocol

Fine-tuned on AGILE, IDH-inhibitor, hypomethylating-agent, and standard 7+3 induction protocols. Proposals come grounded in the regimen the patient is already on.

04

You stay the physician

The agent suggests. Nothing routes anywhere without your signature. Every recommendation is a draft, not a directive.

Interpret Jake Anonymous · AML · Cycle 9 review
Reading 14 panels
SURFACED · AGILE CR criteria 3 citations

ANC has held above the 1.0 K/µL AGILE threshold for 6 consecutive draws (3/3, 3/17, 3/24, 3/31, 4/14, 4/28). However, PLT 67 K/µL and HGB 8.3 g/dL remain below criteria — formal CR will need bone-marrow blast confirmation.

FLAG · trend change 2 citations

HGB shows a −1.8 g/dL drift since C7 start (Feb 1) despite ongoing Aranesp support (3 doses since 2/16). Pattern matches treatment-related cytopenia rather than disease progression.

PROPOSED ACTION draft order

Consider scheduling BM biopsy alongside the C10 plan window (week of 5/19) to capture blast % for AGILE CR confirmation. Continue Ivosidenib at current dose; defer Olutasidenib transition pending biopsy.

What did blasts trend look like vs the 7+3 cohort?

Specialty fit

Built for the panels you actually order.

TrendHUD's longitudinal canvas is configured per specialty out of the box. The metrics that matter, ordered the way you read.

Hematology / Oncology

AML and MDS trajectories across cytopenias, blasts, transfusion dependence. The trends that matter for treatment decisions, visible at a glance.

CBC w/ DiffBlasts %ReticulocytesTransfusion log

Endocrinology

A1C, TSH, free T4, B12, vitamin D — every endocrine panel on a single longitudinal canvas. Spot drift before the patient does.

A1CTSHFree T4B12Vit D

Plastic Surgery

Pre-op labs, intra-op flowsheet values, post-op recovery markers — one continuous chart instead of three siloed systems.

Pre-op CBCFlowsheetPost-op CMPCoagulation
Security & data

Built on standards your compliance team already knows.

FHIR for ingestion, LOINC for normalization, BAA on the enterprise tier. No new vocabulary. No proprietary lock-in.

FHIR-native ingestion

Reads US Core R4 Observation, DiagnosticReport, Encounter resources straight from your EHR export.

LOINC-coded normalization

Every metric mapped to its LOINC code at ingestion. Cross-EHR comparisons stop being a chart-review project.

HIPAA roadmap · BAA on enterprise

Pilot deployments use de-identified exports. Enterprise tier ships with executed BAA and audit logging.

Request access

We're onboarding pilot practices now.

Heme/onc, endocrinology, and plastic surgery are first in line. Tell us your panel mix and we'll get you on a call this week.

3
Active specialties
FHIR R4
Ingestion standard
~18 mo
Avg history rendered
<5s
Time-to-first-chart
Required
Valid work email required
Required
Required
Required