From 4b2f1f23ca516bc51771812b8ca97b8159ceaed9 Mon Sep 17 00:00:00 2001
From: Theodore Li <theo@sim.ai>
Date: Mon, 29 Jun 2026 10:56:52 -0700
Subject: [PATCH] feat(pii): add redaction timing metrics across sidecar and
 persist path

- Log per-request duration in the Presidio sidecar (/analyze, /anonymize)
- Add durationMs to the mask-batch endpoint log line
- Emit per-execution PII redaction timing (stringCount, totalBytes, durationMs, scrubbed)
---
 apps/pii/server.py                            | 20 +++++++++++++++++++
 .../app/api/guardrails/mask-batch/route.ts    |  6 +++++-
 apps/sim/lib/logs/execution/pii-redaction.ts  | 11 ++++++++++
 3 files changed, 36 insertions(+), 1 deletion(-)

diff --git a/apps/pii/server.py b/apps/pii/server.py
index 597fe8f3d90..3fbd9859e45 100644
--- a/apps/pii/server.py
+++ b/apps/pii/server.py
@@ -5,6 +5,8 @@
 endpoints so a single PRESIDIO_URL serves both.
 """
 
+import logging
+import time
 from typing import Any
 
 from fastapi import FastAPI
@@ -133,6 +135,9 @@ def build_analyzer() -> AnalyzerEngine:
 analyzer = build_analyzer()
 anonymizer = AnonymizerEngine()
 
+# Propagates to uvicorn's root handler, so timing lands in the container log stream.
+logger = logging.getLogger("sim.pii")
+
 app = FastAPI(title="Sim Presidio", docs_url=None, redoc_url=None)
 
 
@@ -163,6 +168,7 @@ def supported_entities(language: str = "en") -> list[str]:
 
 @app.post("/analyze")
 def analyze(req: AnalyzeRequest) -> list[dict[str, Any]]:
+    started = time.perf_counter()
     results = analyzer.analyze(
         text=req.text,
         language=req.language,
@@ -170,11 +176,19 @@ def analyze(req: AnalyzeRequest) -> list[dict[str, Any]]:
         score_threshold=req.score_threshold,
         return_decision_process=req.return_decision_process,
     )
+    logger.info(
+        "analyze lang=%s chars=%d entities=%d duration_ms=%.1f",
+        req.language,
+        len(req.text),
+        len(results),
+        (time.perf_counter() - started) * 1000,
+    )
     return [r.to_dict() for r in results]
 
 
 @app.post("/anonymize")
 def anonymize(req: AnonymizeRequest) -> dict[str, Any]:
+    started = time.perf_counter()
     analyzer_results = [
         RecognizerResult(
             entity_type=r["entity_type"],
@@ -197,6 +211,12 @@ def anonymize(req: AnonymizeRequest) -> dict[str, Any]:
         analyzer_results=analyzer_results,
         operators=operators,
     )
+    logger.info(
+        "anonymize chars=%d spans=%d duration_ms=%.1f",
+        len(req.text),
+        len(analyzer_results),
+        (time.perf_counter() - started) * 1000,
+    )
     return {
         "text": result.text,
         "items": [
diff --git a/apps/sim/app/api/guardrails/mask-batch/route.ts b/apps/sim/app/api/guardrails/mask-batch/route.ts
index 696b69e749c..b04d3d21106 100644
--- a/apps/sim/app/api/guardrails/mask-batch/route.ts
+++ b/apps/sim/app/api/guardrails/mask-batch/route.ts
@@ -27,8 +27,12 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
   const { texts, entityTypes, language } = parsed.data.body
 
   try {
+    const startedAt = performance.now()
     const masked = await maskPIIBatch(texts, entityTypes, language)
-    logger.info('Masked PII batch', { count: texts.length })
+    logger.info('Masked PII batch', {
+      count: texts.length,
+      durationMs: Math.round(performance.now() - startedAt),
+    })
     return NextResponse.json({ masked })
   } catch (error) {
     // An unreachable/misconfigured Presidio sidecar makes maskPIIBatch throw; fail
diff --git a/apps/sim/lib/logs/execution/pii-redaction.ts b/apps/sim/lib/logs/execution/pii-redaction.ts
index 8cd0fac5326..5b17694090a 100644
--- a/apps/sim/lib/logs/execution/pii-redaction.ts
+++ b/apps/sim/lib/logs/execution/pii-redaction.ts
@@ -132,6 +132,7 @@ export async function redactPIIFromExecution(
 ): Promise<RedactablePayload> {
   const { entityTypes } = options
   const language = options.language ?? 'en'
+  const startedAt = performance.now()
 
   const units = REDACTABLE_KEYS.filter((key) => payload[key] !== undefined).map((key) => ({
     key,
@@ -151,12 +152,14 @@ export async function redactPIIFromExecution(
   if (collected.length === 0) return payload
 
   let masked: string[]
+  let scrubbed = false
   if (totalBytes > PII_MAX_TOTAL_BYTES) {
     logger.warn('Execution exceeds PII redaction ceiling; scrubbing text', {
       totalBytes,
       ceiling: PII_MAX_TOTAL_BYTES,
     })
     masked = collected.map(() => REDACTION_FAILED_MARKER)
+    scrubbed = true
   } else {
     try {
       // Presidio runs only in the app container; the persist path also runs in
@@ -168,6 +171,7 @@ export async function redactPIIFromExecution(
         stringCount: collected.length,
       })
       masked = collected.map(() => REDACTION_FAILED_MARKER)
+      scrubbed = true
     }
   }
 
@@ -176,5 +180,12 @@ export async function redactPIIFromExecution(
   for (const unit of units) {
     result[unit.key] = transformUnit(unit.key, unit.value, () => masked[index++])
   }
+
+  logger.info('PII redaction completed', {
+    stringCount: collected.length,
+    totalBytes,
+    durationMs: Math.round(performance.now() - startedAt),
+    scrubbed,
+  })
   return result
 }