Home / Prior-art search that shows its mapping
A prior-art search is only as good as its mapping: which passage anticipates which limitation, and what is still missing. SallyIP structures every search as a reviewable claim-to-evidence matrix.
Retrieval breadth, paraphrase-robust matching, and structured matrices a practitioner reviews — including the invented-step combinations the search surfaces for obviousness analysis. Final patentability judgment stays with counsel.
Searches run across EPO, USPTO and EUIPO official sources plus harvested literature, with provider status reported honestly — when a source is unreachable, the run says so instead of simulating results. Every official-search run is persisted with its audit trail, so a search can be re-run and its coverage challenged later.
Each matrix row pairs one claim limitation with the passages that disclose it, the passages that come close, and the limitations with nothing found. Unmapped limitations are findings in their own right: they narrow where the novelty lives and where drafting should reinforce support. Gaps are never filled with model knowledge.
An anticipation found during prosecution costs an amendment; the same reference found in litigation can cost the patent. Search quality compounds: the matrix from this search becomes the novelty argument, the FTO baseline, and the prosecution record. That is why every row links to its passage — a mapping nobody can inspect is a conclusion nobody should rely on.
Run a prior-art checkPatent AI hub
FTO vs patentability · Prior art · Claim charts · retrieval benchmark
Tools that retrieve candidate references, map each claim element to passages, track priority dates and families, and leave gaps explicit. SallyIP adds exact-quote verification so every mapped passage can be inspected word-for-word.