We recorded an elevated number of HTTP 500-level response codes as a result of an increased number of active connections to our querying cluster. Clients may have also experienced higher latencies when accessing the API. Cluster resources have been upscaled, and the issue has been resolved.