diff --git a/bex/ensemble.py b/bex/ensemble.py
index 2ba14d5..0e557f4 100644
--- a/bex/ensemble.py
+++ b/bex/ensemble.py
@@ -386,18 +386,18 @@ _ALGORITHMS = {
 }
 
 
-def _run_kore(sequences, kmax, N):
+def _run_kore(sequences, kmax, N, n_workers=1):
     """Run kOREInference, return (grammar, score) or (None, inf)."""
     from .kore import kOREInference
     kore = kOREInference(k_max=kmax, N=N)
-    result = kore.infer(sequences)
+    result = kore.infer(sequences, n_workers=n_workers)
     if result:
         _, expr, _ = result
         return expr, mdl_score_simple(expr, sequences)
     return None, float('inf')
 
 
-def infer_ensemble(sequences, kmax=2, N=3, prefer=None, min_coverage=1.0, include_kore=False):
+def infer_ensemble(sequences, kmax=2, N=3, prefer=None, min_coverage=1.0, include_kore=False, n_workers=1):
     """Run all applicable algorithms and return the best by MDL score.
 
     Args:
@@ -450,7 +450,7 @@ def infer_ensemble(sequences, kmax=2, N=3, prefer=None, min_coverage=1.0, includ
 
     # 3. kOREInference (opt-in via include_kore=True)
     if include_kore:
-        kore_g, kore_score = _run_kore(sequences, kmax, N)
+        kore_g, kore_score = _run_kore(sequences, kmax, N, n_workers=n_workers)
         if kore_g:
             results.append(('kOREInference', kore_g, kore_score))
 
diff --git a/bex/kore.py b/bex/kore.py
index c960d22..b2e1f6c 100644
--- a/bex/kore.py
+++ b/bex/kore.py
@@ -13,6 +13,9 @@ Unlike the PTA→Shrink→Repair approach from Bex 2008, this follows
 the journal paper (arXiv 1004.2372) exactly.
 """
 
+import os
+from concurrent.futures import ProcessPoolExecutor, as_completed
+
 from .ikoa import ikoa
 from .rwrsq import rwr_sq
 from .idregex import is_deterministic
@@ -53,6 +56,28 @@ def validate_k_ore(expr, k, alphabet_set=None):
     return True, "OK"
 
 
+def _kore_trial(args):
+    """Run one (k, n) trial. Module-level for ProcessPoolExecutor.
+
+    Args:
+        args: (sequences, k, trial_index) — trial_index unused, for diagnostics
+
+    Returns:
+        (koa_automaton, expression_string, k) or None
+    """
+    sequences, k, _ = args
+    G = ikoa(sequences, k, num_trials=1)
+    if G is None:
+        return None
+    expr = rwr_sq(G)
+    if expr and expr not in ('∅', 'ε'):
+        if is_deterministic(expr):
+            valid, _ = validate_k_ore(expr, k)
+            if valid:
+                return (G, expr, k)
+    return None
+
+
 class kOREInference:
     """
     |———— Algorithm 4: iDRegEx ————|
@@ -72,9 +97,12 @@ class kOREInference:
         self.k_max = k_max
         self.N = N
 
-    def infer(self, sequences):
-        """
-        Infer the best k-ORE for the given sequences.
+    def infer(self, sequences, n_workers=1):
+        """Infer the best k-ORE for the given sequences.
+
+        Args:
+            sequences: list of token sequences
+            n_workers: parallel workers. >1 runs (k, n) trials concurrently.
 
         Returns:
             (koa_automaton, expression_string, best_k) or None if no valid
@@ -84,19 +112,25 @@ class kOREInference:
         if not sequences:
             return None
 
+        trials = [(sequences, k, i)
+                  for k in range(1, self.k_max + 1)
+                  for i in range(self.N)]
+
         candidates = []
 
-        for k in range(1, self.k_max + 1):
-            for _ in range(self.N):
-                G = ikoa(sequences, k, num_trials=1)
-                if G is None:
-                    continue
-                expr = rwr_sq(G)
-                if expr and expr not in ('∅', 'ε'):
-                    if is_deterministic(expr):
-                        valid, _ = validate_k_ore(expr, k)
-                        if valid:
-                            candidates.append((G, expr, k))
+        if n_workers <= 1 or len(trials) < 2:
+            for t in trials:
+                result = _kore_trial(t)
+                if result is not None:
+                    candidates.append(result)
+        else:
+            nw = min(n_workers, len(trials))
+            with ProcessPoolExecutor(max_workers=nw) as ex:
+                futures = [ex.submit(_kore_trial, t) for t in trials]
+                for f in as_completed(futures):
+                    result = f.result()
+                    if result is not None:
+                        candidates.append(result)
 
         if not candidates:
             return None