One running record of your bloodwork. None of the paperwork it came in.
PanelShift turns every Quest or LabCorp PDF into a structured row in your marker history — every value, every reference range, every flag, across every draw. The identifying paperwork stays out of the pipeline. Informational tracking — not a diagnostic tool.
Private beta is invite-only. We onboard in small batches to keep extraction quality, privacy controls, and support tight.
LAB #: 24-█████████
Approved beta users can lock founding rates for 12 months.
For approved beta users, applied automatically at checkout. Founding rate persists for the first 12 months, then graduates to the public rate.
Most of your health data is a guess. Bloodwork is a measurement.
A single panel reads hormones, lipids, immune signals, organ function, and metabolic state directly from your blood. It surfaces silent shifts early, quantifies risk in numbers instead of vibes, informs the conversations you have with your clinician, and shows whether what you're actually doing is moving the dial — or not.
- 01Source data, not inference
Wearables estimate sleep, recovery, and effort. A blood panel measures hormones, lipids, glucose, and organ function directly.
- 02Baseline, then repeat
A single reading tells you very little. The same key markers run consistently against your last five draws tells you a lot.
- 03Read against your history
Numbers gain meaning against your prior values, your symptoms, and the conversations you have with your clinician.
Upload. Extract. Stack against every previous draw.
Most lab reports live in email attachments and patient portals nobody revisits. PanelShift converts every upload into a structured, comparable record — and discards the PDF when it's done.
Drop a PDF
Quest or LabCorp report, up to 40 pages. The buffer lives in memory only — nothing is written to disk at any point.
Markers extracted, identifiers left behind
Name, DOB, MRN, provider NPI, address, and barcodes are redacted before the extraction model ever runs. What we keep is a structured marker set — values, units, reference ranges, flags. The PDF is discarded immediately after.
AI flags deltas, writes the summary
Every value is parsed with its reference range, cross-referenced against your previous draws, and summarized for trend context and out-of-range flags. Informational only — not diagnosis or treatment guidance.
Your labs, your card, and your account live in three different systems. None of them know about the others.
This isn't a privacy policy — it's the product architecture. Stripe holds your card. Clerk holds your session. PanelShift holds a marker history with no name, DOB, or address attached to it.
Identifiers stripped before the AI sees anything
A dedicated redaction pipeline runs on raw PDF text before extraction starts. Name, DOB, physician name, NPI, accession numbers, and addresses are removed before AI processing and before structured marker data is stored.
The PDF never touches disk
Your file exists in memory only for the duration of the extraction pipeline. Raw PDFs are not persisted; transient processing buffers and extracted text are discarded after structured markers are written.
The analysis model sees markers, not a patient
The analysis model receives marker values, units, reference ranges, flags, and a draw date. No name, DOB, address, physician, accession number, or billing context is included in the payload.
We never see your card
Payments are processed by Stripe. Card details are collected and stored by Stripe — not PanelShift. We store only the customer, subscription, and status identifiers needed to manage billing. No card number or payment credentials touch our servers.
The product is the marker history — not the PDF.
Every upload adds a row to your record. Open any marker to see its full series — values, reference ranges, and flags across every draw you've uploaded.
| Marker | Value | Ref | Trend | Flag |
|---|---|---|---|---|
| LDL Cholesterol | 128 mg/dL | < 99 | ▲ High | |
| Apolipoprotein B | 104 mg/dL | < 90 | ▲ High | |
| 25-Hydroxy Vitamin D | 27 ng/mL | 30–100 | ▼ Low | |
| Hemoglobin A1c | 5.4 % | 4–5.6 | In range | |
| Glucose, Fasting | 94 mg/dL | 65–99 | In range |
LDL Cholesterol at 128 mg/dL — above the listed reference target and up from 106 across the prior draws. ApoB moved in the same direction at 104 mg/dL.
Vitamin D is below range at 27 ng/mL. That low flag has persisted across the last three synthetic draws.
A1c, fasting glucose, kidney markers, liver enzymes, and thyroid markers are stable relative to recent history.
Built around the marker. Not the report.
Longitudinal by default
Every draw becomes a structured row. Every value lands next to your prior values for the same marker. No concept of "current report" vs "old report" — there's one history.
Quest & LabCorp native
Extraction tuned to both lab report layouts. Reflex tests, calculated values, sensitivity assays, and multi-panel reports all parse — not just the headline panel.
Reference-range aware
The lab's stated reference range is captured per draw, because it drifts between labs and over time. In-range at Quest isn't always in-range at LabCorp — your history reflects that.
Plain-English informational analysis
Every upload triggers a write-up against your last five draws. Deltas, trends, and out-of-range flags are called out directly — no diagnosis, no treatment guidance.
Optional context notes
Add your own notes to any draw — context the lab report doesn't know about, like supplements, meds, or recent changes. Analysis reads them as context for marker trends, never as a source of truth.
Marker trend view
Click any marker to open its full series — every value, every reference range, every flag, across every draw. Sparklines on the dashboard click into the same view.
Pricing
Start with the latest panel. Pay when history matters.
Private beta access is waitlist-based. Pricing shown is what approved beta users can choose after onboarding: Free explains your current report, Pro adds history, and Advanced adds protocol context.
Best for trying one report
Free
$0
Use PanelShift on your latest bloodwork before you pay for longitudinal tracking.
- 3 lifetime uploads for digital Quest or LabCorp PDFs
- Current-panel AI summary in plain English
- Marker table with values, units, ranges, and lab flags
- Private extraction flow: identifiers are stripped and the raw PDF is not stored
- Good for deciding whether the product is useful on one current report
Best for recurring bloodwork
Pro
$10/mo
$15/mo
Founding Member · First 12 months
For people who run labs repeatedly and want the current draw interpreted against prior results.
Monthly billing in USD.
- Unlimited uploads
- Full marker history across stored draws
- Trend charts for seeing what changed over time
- AI analysis using historical context from up to the last 5 draws
- Monthly AI chat allowance for follow-up questions about your labs
Best for protocol context
Advanced
$19/mo
$29/mo
Founding Member · First 12 months
For TRT, performance, and high-frequency users comparing marker shifts against self-entered context.
Monthly billing in USD.
- Everything in Pro
- Medication, supplement, and protocol context logging
- Protocol-aware upload analysis and chat context
- 2× Pro's monthly AI allowance for heavier compare and trend questions
- Useful when you need to line up lab changes with protocol start, stop, or dose changes
Monthly billing in USD through Stripe. Cancel in Settings; paid access continues through the current billing period. See the refund policy and terms.
You already have bloodwork in your email. Make it a record.
Private beta is invite-only. We onboard in small batches to keep extraction quality, privacy controls, and support tight.