Live demoSample data, real AI.
The charts below are illustrative filler, not the real corpus, but the assistant is fully live and runs instantly: no setup, no API key, no sign-up.
This is the analytics dashboard + the grounded "Ask the corpus" assistant, running on placeholder numbers. On the real product these are computed from every published BVA decision (1992–present) and every statute/regulation citation is verified against the dated Title 38 store.
sample
Decisions
1,000,000
illustrative
sample
Grant rate
31.2%
≥1 issue granted
sample
Remand rate
22.4%
sent back
sample
Deny rate
25.1%
fully denied
sample
AMA share
17.0%
post-2019
Disposition mix sample
Top conditions · grant rate sample
Ask the corpus live AIBuilt-in model · no key needed
3 free AI analyses per visitor, no sign-up.
Grounded, not guessing. Every statistic is locked to the data block; the model can't invent a number.
Citations verified. Each 38 U.S.C./CFR cite is resolved against the dated Title 38 store; unverified cites are struck through.
Nothing to set up. Runs instantly on VALORDB's built-in model, no API key, no install, no sign-up. Free for veterans.
Attorney workflows (Pro). Brief-Ready paragraph, Distinguishing rebuttal, and Memo-Outline skeletons, plus a bring-your-own-key / local-Ollama sovereign lane for privileged work.
A look inside VALORDB
Live views of the product on the real 1.86-million-decision corpus, the outcome dashboard, the Explore breakdowns, the verbatim Title 38 reader, and grounded AI. Aggregate figures only; no veteran is identified.
Outcome dashboard. Grant, remand & denial rates across 1.86M decisions, with year trends and disposition mix.
Explore. Slice by representation, condition, Regional Office or judge, every rate shows its N and its gap from the national average.
The complete Title 38. Every statute & regulation, verbatim and dated, full-text searchable with cross-references linked.
Ask the corpus. Grounded answers with every statute citation checked against the dated Title 38 store.