perf(ai-sdk): embed query once across router fan-out + fold umlauts in intent/concept matching
CI / detect-changes (push) Successful in 5s
CI / branch-name (push) Has been skipped
CI / guardrail-integrity (push) Has been skipped
CI / secret-scan (push) Has been skipped
CI / dep-audit (push) Has been skipped
CI / sbom-scan (push) Has been skipped
CI / build-sha-integrity (push) Successful in 5s
CI / validate-canonical-controls (push) Successful in 4s
CI / loc-budget (push) Successful in 18s
CI / go-lint (push) Has been skipped
CI / python-lint (push) Has been skipped
CI / nodejs-lint (push) Has been skipped
CI / nodejs-build (push) Successful in 3m0s
CI / test-go (push) Successful in 59s
CI / iace-gt-coverage (push) Successful in 17s
CI / test-python-backend (push) Has been skipped
CI / test-python-document-crawler (push) Has been skipped
CI / test-python-dsms-gateway (push) Has been skipped

Authority Router re-embedded the query per collection (6x); on dev the embed
endpoint (OVH) is remote so that was 6 round-trips = 7-12s per /retrieve. Embed
once, reuse via ctx across the concurrent per-collection searches.
DetectIntent + ConceptNorms now fold ae/oe/ue/ss so ASCII (Pruefe) and umlaut
(Pruefe) inputs both match.
This commit is contained in:
Claude
2026-07-01 19:03:11 +02:00
parent cf2cea437e
commit 0903e3a8d1
7 changed files with 87 additions and 5 deletions
@@ -48,12 +48,12 @@ var legalConceptOntology = []conceptNorm{
// ConceptNorms returns the load-bearing norm_ids for the concepts named in the
// query (dedup, order-preserving). Empty if no concept is named.
func ConceptNorms(query string) []string {
q := strings.ToLower(query)
q := normalizeGerman(query)
seen := map[string]bool{}
out := []string{}
for _, cn := range legalConceptOntology {
for _, kw := range cn.keywords {
if strings.Contains(q, kw) {
if strings.Contains(q, normalizeGerman(kw)) {
for _, nid := range cn.normIDs {
if !seen[nid] {
seen[nid] = true