remove ttl for cache and just use maximumSize

Author: ellenmuller

media-api/app/controllers/MediaApi.scala

@@ -56,7 +56,6 @@ class MediaApi(
   // subsequent requests within the TTL window skip Bedrock entirely.
   private val embeddingCache: AsyncLoadingCache[String, List[Float]] = Scaffeine()
     .maximumSize(config.aiSearchEmbeddingCacheMaxSize)
-    .expireAfterWrite(config.aiSearchEmbeddingCacheTtlSeconds.seconds)
     .buildAsyncFuture((normQuery: String) =>
       embedder.createQueryEmbedding(normQuery)(MarkerMap())
     )
@@ -684,7 +683,6 @@ class MediaApi(
     }
 
     SearchParams.validate(searchParams).fold(
-      // TODO: respondErrorCollection?
       errors => Future.successful(respondError(UnprocessableEntity, InvalidUriParams.errorKey,
         errors.map(_.message).mkString(", "))
       ),

media-api/app/lib/MediaApiConfig.scala

@@ -79,11 +79,7 @@ class MediaApiConfig(resources: GridConfigResources) extends CommonConfigWithEla
   // Semantic search retrieves from a bounded nearest-neighbour window.
   // This keeps KNN query cost predictable and makes result totals explicitly capped.
   val aiSearchSemanticRetrievalWindow: Int = intOpt("ai.search.semanticRetrievalWindow").getOrElse(1000)
-
-  // Embedding cache: avoids repeated Bedrock calls for the same text query within the TTL window.
-  // In-flight requests for the same key are also deduplicated (only one Bedrock call fires).
   val aiSearchEmbeddingCacheMaxSize: Int = intOpt("ai.search.embeddingCache.maxSize").getOrElse(500)
-  val aiSearchEmbeddingCacheTtlSeconds: Int = intOpt("ai.search.embeddingCache.ttlSeconds").getOrElse(600)
 
   val maybeAgencyPickQuery: Option[Query] = agencyPicksIngredients.map { ingredients =>
     filters.or(