diff --git a/core/nurse_scheduling/scheduler.py b/core/nurse_scheduling/scheduler.py
index 06cf8da4..db29759d 100644
--- a/core/nurse_scheduling/scheduler.py
+++ b/core/nurse_scheduling/scheduler.py
@@ -49,9 +49,11 @@ def schedule(
     ctx.n_people = len(ctx.people.items)
     ctx.dates.items = [ctx.dates.range.startDate + timedelta(days=d) for d in range(ctx.n_days)]
 
-    # Map shift type ID to shift type index
+    # Map shift type ID to shift type index.
+    # IDs are stringified so unquoted YAML scalars (e.g. `0` parsed as int)
+    # produce the same map key as the string lookups in `parse_sids`.
     for s in range(ctx.n_shift_types):
-        ctx.map_sid_s[ctx.shiftTypes.items[s].id] = [s]
+        ctx.map_sid_s[str(ctx.shiftTypes.items[s].id)] = [s]
     # Add shift type ALL and OFF keywords
     ctx.map_sid_s[ALL] = list(range(ctx.n_shift_types))
     ctx.map_sid_s[OFF] = [OFF_sid]
@@ -59,17 +61,18 @@ def schedule(
     for g in range(len(ctx.shiftTypes.groups)):
         group = ctx.shiftTypes.groups[g]
         # Flatten and deduplicate shift type indices for the group
-        ctx.map_sid_s[group.id] = sorted(set().union(*[ctx.map_sid_s[sid] for sid in group.members]))
-    # Map person ID to person index
+        ctx.map_sid_s[str(group.id)] = sorted(set().union(*[ctx.map_sid_s[str(sid)] for sid in group.members]))
+    # Map person ID to person index.
+    # IDs are stringified for the same reason as shift type IDs above.
     for p in range(ctx.n_people):
-        ctx.map_pid_p[ctx.people.items[p].id] = [p]
+        ctx.map_pid_p[str(ctx.people.items[p].id)] = [p]
     # Add people ALL keyword
     ctx.map_pid_p[ALL] = list(range(ctx.n_people))
     # Map people group ID to list of person indices
     for g in range(len(ctx.people.groups)):
         group = ctx.people.groups[g]
         # Flatten and deduplicate person indices for the group
-        ctx.map_pid_p[group.id] = sorted(set().union(*[ctx.map_pid_p[pid] for pid in group.members]))
+        ctx.map_pid_p[str(group.id)] = sorted(set().union(*[ctx.map_pid_p[str(pid)] for pid in group.members]))
 
     # Map date string (YYYY-MM-DD) to date index
     if ctx.country is not None and ctx.country != 'TW':
diff --git a/core/nurse_scheduling/utils.py b/core/nurse_scheduling/utils.py
index 38f6a10d..9723b7aa 100644
--- a/core/nurse_scheduling/utils.py
+++ b/core/nurse_scheduling/utils.py
@@ -88,7 +88,9 @@ def parse_dates(dates, map_did_d, date_range):
     return sorted(set(result))
 
 def parse_sids(sids, map_sid_s):
-    sids = ensure_list(sids)
+    # Coerce to str so that unquoted YAML scalars (e.g. `0` parsed as int)
+    # match the string-keyed `map_sid_s`. Mirrors `parse_dates`'s `map(str, ...)`.
+    sids = [str(sid) for sid in ensure_list(sids)]
     result = []
     for sid in sids:
         if sid not in map_sid_s:
@@ -97,7 +99,9 @@ def parse_sids(sids, map_sid_s):
     return sorted(set(result))
 
 def parse_pids(pids, map_pid_p):
-    pids = ensure_list(pids)
+    # Coerce to str so that unquoted YAML scalars (e.g. `0` parsed as int)
+    # match the string-keyed `map_pid_p`. Mirrors `parse_dates`'s `map(str, ...)`.
+    pids = [str(pid) for pid in ensure_list(pids)]
     result = []
     for pid in pids:
         if pid not in map_pid_p: