diff --git a/.env.example b/.env.example
new file mode 100644
index 0000000..e081a81
--- /dev/null
+++ b/.env.example
@@ -0,0 +1,13 @@
+# === Azure OpenAI ===
+AZURE_OPENAI_ENDPOINT=
+AZURE_OPENAI_API_KEY=
+AZURE_OPENAI_API_VERSION=2024-05-01-preview
+AZURE_OPENAI_DEPLOYMENT=
+
+# === Azure AI Search ===
+AZURE_SEARCH_ENDPOINT=
+AZURE_SEARCH_API_KEY=
+AZURE_SEARCH_INDEX_NAME=
+
+# Optional local API for eval (POST /ask)
+RAG_API_URL=http://localhost:8000/ask
diff --git a/tools/eval.py b/tools/eval.py
new file mode 100644
index 0000000..d64f8ca
--- /dev/null
+++ b/tools/eval.py
@@ -0,0 +1,53 @@
+import os, csv, argparse, json, time
+
+RAG_API_URL = os.getenv("RAG_API_URL", "http://localhost:8000/ask")
+
+def _ask_via_rest(q: str) -> str:
+    import requests
+    r = requests.post(RAG_API_URL, json={"question": q}, timeout=60)
+    r.raise_for_status()
+    data = r.json()
+    return data.get("answer") or data.get("output_text") or json.dumps(data)
+
+def normalize(s: str) -> str: return (s or "").strip().lower()
+def exact_match(p: str, g: str) -> float: return 1.0 if normalize(p)==normalize(g) and g else 0.0
+
+def token_f1(p: str, g: str) -> float:
+    P, G = normalize(p).split(), normalize(g).split()
+    if not P and not G: return 1.0
+    if not P or not G:  return 0.0
+    common = set(P)&set(G)
+    if not common: return 0.0
+    prec = sum(w in G for w in P)/len(P)
+    rec  = sum(w in P for w in G)/len(G)
+    return 0.0 if prec+rec==0 else 2*prec*rec/(prec+rec)
+
+def run_eval(rows):
+    out = []
+    for q,g in rows:
+        try: pred = _ask_via_rest(q)
+        except Exception as e: pred = f"[ERROR] {e}"
+        em = exact_match(pred,g) if g else ""
+        f1 = token_f1(pred,g) if g else ""
+        print(f"Q: {q}
+A: {pred}
+EM: {em}  F1: {f1}
+---")
+        out.append((q,pred,g,em,f1))
+    with open("results.csv","w",newline="",encoding="utf-8") as f:
+        w=csv.writer(f); w.writerow(["question","pred","gold","exact_match","f1"]); w.writerows(out)
+    print(f"Saved results.csv (n={len(out)})")
+
+def main():
+    ap=argparse.ArgumentParser()
+    ap.add_argument("--q", type=str)
+    ap.add_argument("--file", type=str)
+    a=ap.parse_args()
+    if a.q: run_eval([(a.q,"")])
+    elif a.file:
+        with open(a.file,newline="",encoding="utf-8") as f:
+            r=csv.DictReader(f)
+            rows=[(d["question"], d.get("answer","")) for d in r]
+        run_eval(rows)
+    else: ap.error("Provide --q or --file")
+if __name__=="__main__": main()
diff --git a/tools/eval_qna.csv b/tools/eval_qna.csv
new file mode 100644
index 0000000..6f29ccf
--- /dev/null
+++ b/tools/eval_qna.csv
@@ -0,0 +1,3 @@
+question,answer
+What is your return policy?,Returns accepted within 30 days with receipt.
+How do I reset my password?,Use the “Forgot Password” link and follow the email instructions.