Clinical timeline
Every uploaded document on a single chronological spine, with source language flags.
Specialist-intake workflow
Turn fragmented medical records into a specialist-ready second-opinion packet.
Upload PDFs from different hospitals and languages. Get a safe, structured PDF handoff with timeline, key findings, missing records, and source provenance.
Built for visit preparation. Not diagnosis. Not treatment advice.
Second-opinion packet
PDF · A4Case A-001 · Nephrology
3 documents · 2022-03-01 → 2024-01-01 · 🇮🇹 🇩🇪 🇬🇧
Active problems
Persistently flagged Creatinine across reports
What changed
Creatinine 1.2 → 1.5 → 1.8 mg/dL · upward trend
Missing for nephrology
Recent urinalysis · Kidney imaging · BP record
Questions for the specialist
What additional renal workup is appropriate based on these records?
From scattered records to handoff
One workflow, four steps.
No dashboard. No chat. A linear path from messy uploads to a clinician-ready PDF.
1Upload records
Drop PDFs or click to browse
- 📄 cmp_2022.pdf
- 📄 ecografia_2023.pdf
- 📄 renal_panel_2024.pdf
2Completeness check
Coverage72%
- ✓ Lab report
- ✓ Imaging
- ✗ Urinalysis
3Packet preview
Active problems
Persistently flagged Creatinine across reports
Key findings
Creatinine 1.8 mg/dL [HIGH]
Provenance
3 sources cited
4Download PDF
second_opinion_packet.pdf
A4 · multilingual · provenance attached
How it works
Predictable, deterministic, four steps.
- 01
Upload records
Drop in PDFs from any combination of hospitals, languages, and visits. Lab reports, imaging, clinical notes, pathology, discharge summaries.
- 02
Choose specialty & language
Pick the receiving specialty (nephrology, cardiology, oncology, …) and the output language. Internal canonical names stay English.
- 03
Review what's missing
A specialty-aware completeness card surfaces missing categories, priority metrics, and recommended uploads — before the LLM call.
- 04
Download doctor-ready PDF
Real PDF (not HTML), with cover page, clinical timeline, key findings, missing-info checklist, questions, and provenance appendix.
What the packet includes
Nine sections built for clinical handoff.
Every section is provenance-backed and deterministic-first. The LLM phrases — it does not invent priority, structure, or trends.
Key findings
Flagged labs, imaging impressions, pathology highlights — grouped by category.
What changed over time
Cross-document trend detection on canonical metrics with deterministic significance.
Missing-records checklist
Specialty-aware. Critical vs recommended. Tells the patient what to ask for next.
Questions for the specialist
Bounded prompts grounded in the records — designed to be answerable in under a minute.
Source provenance
Every claim cites the exact document, date, and language it came from. Always.
Why it matters
Fragmented records become a usable intake packet.
Specialists don't want 30 raw PDFs.
They want one structured packet they can read in two minutes — timeline, key findings, sources. The product produces exactly that.
Patients don't know what's missing.
A specialty-aware completeness check shows what categories the receiving specialist usually expects, before the case is sent.
Multilingual records make handoff harder.
Per-document language detection keeps source records intact; the output packet is generated in one language of your choice.
Where we are
Honest, early, working.
No invented testimonials. No vanity metrics. The product is in real-case validation right now — these placeholders update as evidence lands.
- In progress
Testing with 10 patient cases
- Pending
Clinician review in progress
- Built from
Real fragmented medical-record workflows
Generate your second-opinion packet.
Upload PDFs, review what's missing, hand the specialist a structured PDF — in the language they speak.
Built for visit preparation. Not diagnosis. Not treatment advice.