Every favorable and unfavorable fact, surfaced — the call stays yours.
Case strength and weakness analysis software from Medrecords AI flags every favorable and unfavorable fact directly from the record — pre-existing conditions, treatment gaps, corroborating imaging, inconsistent complaints — each cited to the page. The AI surfaces the signal; no merit score is assigned or implied, and the call stays with you.
Pre-existing conditions, found before the defense finds them.
The AI reads the whole record, not just the post-incident pages, and flags earlier treatment, complaints, or imaging involving the same body part — dated, attributed, and cited. Whether a prior entry hurts or is distinguishable is your argument to make; the point is that it never surfaces for the first time in opposing counsel's exhibit list.
Gaps and inconsistencies, before deposition finds them.
The claimed injury history is checked against the documented timeline: when complaints first appear, whether treatment runs continuous or gapped, whether the intake story matches what providers wrote down visit by visit. Every mismatch is flagged with the dates and pages that define it.
Signals, not a score — by design.
This feature will not tell you whether to take the case, and it assigns no number that pretends to. It hands you the documented fact patterns, both directions, each one cited — and stops there. A scored case-merit assessment exists separately on our roadmap, gated behind mandatory professional sign-off.
Flagged, cited, defensible — not guessed.
A strategy memo built on unsourced impressions collapses the first time someone checks the record. Every signal here is source-linked and legally defensible: the page, the provider, the passage. What the record doesn't support, the list doesn't say.
See Verifiable AI CitationsFrom records dump to signal list.
Three steps — then the strategy work starts where it should.
The whole record, prior history included — that's where the unfavorable signals hide.
Pre-existing conditions, gaps, corroboration, inconsistencies — each marked favorable or unfavorable, cited.
Verify any flag against its page, then make the intake, strategy, or settlement call — yours alone.
Who works the signal list.
Built for legal teams on either side of the file.
Intake screening and case strategy that starts from the documented facts — both directions, cited.
For law firmsKnow the prior-injury history and the treatment gaps before demand — not at mediation.
For PI practicesCorroborating documentation and timeline inconsistencies surfaced across dense clinical records.
For med-mal teamsSignal flagging, answered.
A signal is a documented fact pattern — a prior condition, a treatment gap, corroborating imaging — flagged from the record with a page citation, marked favorable or unfavorable. A merit score is a judgment about the case as a whole. This feature produces only signals; no merit score is assigned or implied as final, and the case-merit call stays with the attorney. A scored assessment exists separately on our roadmap, gated behind mandatory professional sign-off.
The AI reads the entire record — not just the post-incident pages — and flags earlier treatment, complaints, or imaging involving the same body part or condition, with the date, provider, and page citation for each. Whether a prior entry helps or hurts is argued by you; the feature's job is making sure nothing surfaces for the first time in opposing counsel's exhibit list.
Favorable signals include objective corroboration — imaging findings that match the complaint, consistent documentation across providers, a clean prior history. Unfavorable signals include treatment gaps, inconsistent complaint histories, pre-existing conditions, and provider statements that cut against causation. Both directions are flagged with equal rigor; the same record is read the way opposing counsel will read it.
The claimed injury history is compared against the documented timeline: when complaints first appear, whether treatment is continuous or gapped, and whether the story told at intake matches what providers recorded visit by visit. Gaps and inconsistencies are flagged with the dates and pages that define them, so each one can be verified or explained.
Yes. Every signal — favorable or unfavorable — carries a page-level citation to the record passage behind it, and clicking the flag opens the source. If the record is ambiguous or a page is illegible, that is flagged as such rather than guessed. No signal enters the list without a source you can verify.
Related capabilities.
Signal flagging sits alongside the platform's other legal-review layers.
The scored version — on the roadmap, gated behind mandatory professional sign-off.
See merit assessmentThe documented care matched against published references — context for the signals.
See literature matchingThe records the file refers to but doesn't contain — the gaps behind the gaps.
See missing recordsAmended or conflicting versions of the same note, surfaced and cited.
See alteration detectionSee the signals in one of your own files.
Upload a single file and get the favorable and unfavorable fact patterns back, cited to the page — no score, no verdict. Handled under our BAA; never used to train a model.