diff --git a/passbook/policy/engine.py b/passbook/policy/engine.py
index 7c3af7b4c..341518afb 100644
--- a/passbook/policy/engine.py
+++ b/passbook/policy/engine.py
@@ -1,14 +1,14 @@
 """passbook policy engine"""
 from multiprocessing import Pipe
 from multiprocessing.connection import Connection
-from typing import List, Tuple
+from typing import List, Tuple, Dict
 
 from django.core.cache import cache
 from django.http import HttpRequest
 from structlog import get_logger
 
 from passbook.core.models import Policy, User
-from passbook.policy.struct import PolicyRequest
+from passbook.policy.struct import PolicyRequest, PolicyResult
 from passbook.policy.process import PolicyProcess
 
 LOGGER = get_logger()
@@ -16,6 +16,18 @@ LOGGER = get_logger()
 def _cache_key(policy, user):
     return f"policy_{policy.pk}#{user.pk}"
 
+class PolicyProcessInfo:
+
+    process: PolicyProcess
+    connection: Connection
+    result: PolicyResult = None
+    policy: Policy
+
+    def __init__(self, process: PolicyProcess, connection: Connection, policy: Policy):
+        self.process = process
+        self.connection = connection
+        self.policy = policy
+
 class PolicyEngine:
     """Orchestrate policy checking, launch tasks and return result"""
 
@@ -23,12 +35,13 @@ class PolicyEngine:
     __request: HttpRequest
     __user: User
 
-    __proc_list: List[Tuple[Connection, PolicyProcess]] = []
+    __processes: List[PolicyProcessInfo] = []
 
     def __init__(self, policies, user: User = None, request: HttpRequest = None):
         self.policies = policies
         self.__request = request
         self.__user = user
+        self.__processes = []
 
     def for_user(self, user: User) -> 'PolicyEngine':
         """Check policies for user"""
@@ -57,36 +70,35 @@ class PolicyEngine:
         for policy in self._select_subclasses():
             cached_policy = cache.get(_cache_key(policy, self.__user), None)
             if cached_policy:
-                LOGGER.debug("Taking result from cache", policy=policy.pk.hex)
+                LOGGER.debug("Taking result from cache", policy=policy)
                 cached_policies.append(cached_policy)
             else:
-                LOGGER.debug("Evaluating policy", policy=policy.pk.hex)
+                LOGGER.debug("Evaluating policy", policy=policy)
                 our_end, task_end = Pipe(False)
-                task = PolicyProcess()
-                task.ret = task_end
-                task.request = request
-                task.policy = policy
-                LOGGER.debug("Starting Process", class_name=task.__class__.__name__)
+                task = PolicyProcess(policy, request, task_end)
+                LOGGER.debug("Starting Process", for_policy=policy)
                 task.start()
-                self.__proc_list.append((our_end, task))
+                self.__processes.append(PolicyProcessInfo(process=task,
+                                                           connection=our_end, policy=policy))
         # If all policies are cached, we have an empty list here.
-        if self.__proc_list:
-            for _, running_proc in self.__proc_list:
-                running_proc.join()
+        for proc_info in self.__processes:
+            proc_info.process.join(proc_info.policy.timeout)
+            # Only call .recv() if no result is saved, otherwise we just deadlock here
+            if not proc_info.result:
+                proc_info.result = proc_info.connection.recv()
         return self
 
     @property
     def result(self) -> Tuple[bool, List[str]]:
         """Get policy-checking result"""
         messages: List[str] = []
-        for our_end, _ in self.__proc_list:
-            policy_result = our_end.recv()
+        for proc_info in self.__processes:
             # passing = (policy_action == Policy.ACTION_ALLOW and policy_result) or \
             #           (policy_action == Policy.ACTION_DENY and not policy_result)
-            LOGGER.debug('Result=%r => %r', policy_result, policy_result.passing)
-            if policy_result.messages:
-                messages += policy_result.messages
-            if not policy_result.passing:
+            LOGGER.debug("Result", passing=proc_info.result.passing)
+            if proc_info.result.messages:
+                messages += proc_info.result.messages
+            if not proc_info.result.passing:
                 return False, messages
         return True, messages
 
diff --git a/passbook/policy/process.py b/passbook/policy/process.py
index 8b09d0cf3..9427e93f2 100644
--- a/passbook/policy/process.py
+++ b/passbook/policy/process.py
@@ -17,10 +17,16 @@ def _cache_key(policy, user):
 class PolicyProcess(Process):
     """Evaluate a single policy within a seprate process"""
 
-    ret: Connection
+    connection: Connection
     policy: Policy
     request: PolicyRequest
 
+    def __init__(self, policy: Policy, request: PolicyRequest, connection: Connection):
+        super().__init__()
+        self.policy = policy
+        self.request = request
+        self.connection = connection
+
     def run(self):
         """Task wrapper to run policy checking"""
         LOGGER.debug("Running policy", policy=self.policy,
@@ -38,5 +44,4 @@ class PolicyProcess(Process):
         # cache_key = _cache_key(self.policy, self.request.user)
         # cache.set(cache_key, (self.policy.action, policy_result, message))
         # LOGGER.debug("Cached entry as %s", cache_key)
-        self.ret.send(policy_result)
-        self.ret.close()
+        self.connection.send(policy_result)