diff --git a/src/faceblur/cli.py b/src/faceblur/cli.py
new file mode 100644
index 0000000..0362f0f
--- /dev/null
+++ b/src/faceblur/cli.py
@@ -0,0 +1,108 @@
+"""CLI module for faceblur-poc."""
+
+import argparse
+import sys
+from pathlib import Path
+
+from .video import extract_frames
+from .detect import FaceDetector
+from .cluster import cluster_faces
+from .output import generate_output
+
+
+def main():
+    """Main entry point for CLI."""
+    parser = argparse.ArgumentParser(
+        description="Face detection and clustering POC",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+
+    subparsers = parser.add_subparsers(dest="command", help="Available commands")
+
+    detect_parser = subparsers.add_parser(
+        "detect", help="Detect and cluster faces in video"
+    )
+    detect_parser.add_argument(
+        "--video", required=True, help="Path to input video file"
+    )
+    detect_parser.add_argument("--output", default="output", help="Output directory")
+    detect_parser.add_argument(
+        "--interval", type=int, default=30, help="Frame interval"
+    )
+    detect_parser.add_argument(
+        "--eps", type=float, default=0.4, help="DBSCAN eps parameter (cosine distance)"
+    )
+    detect_parser.add_argument(
+        "--min-samples", type=int, default=2, help="DBSCAN min_samples"
+    )
+
+    args = parser.parse_args()
+
+    if args.command is None:
+        parser.print_help()
+        sys.exit(1)
+
+    if args.command == "detect":
+        run_detect(args)
+
+
+def run_detect(args):
+    """Run the detect command."""
+    print(f"Processing video: {args.video}")
+    print(f"Output directory: {args.output}")
+    print(f"Frame interval: {args.interval}")
+
+    video_path = Path(args.video)
+    if not video_path.exists():
+        print(f"Error: Video file not found: {args.video}", file=sys.stderr)
+        sys.exit(1)
+
+    print("\n[1/5] Extracting frames...")
+    frames = extract_frames(args.video, f"{args.output}/frames_original", args.interval)
+    print(f"Extracted {len(frames)} frames")
+
+    if not frames:
+        print("Error: No frames extracted from video", file=sys.stderr)
+        sys.exit(1)
+
+    print("\n[2/5] Initializing face detector...")
+    detector = FaceDetector()
+
+    print("\n[3/5] Detecting faces...")
+    all_faces = []
+    for i, frame in enumerate(frames):
+        try:
+            faces = detector.detect_faces(frame.path, frame.index)
+            all_faces.extend(faces)
+            print(
+                f"  Frame {i + 1}/{len(frames)}: {len(faces)} faces (total: {len(all_faces)})"
+            )
+        except Exception as e:
+            print(f"  Warning: Failed to detect faces in frame {frame.index}: {e}")
+
+    print(f"Total faces detected: {len(all_faces)}")
+
+    if not all_faces:
+        print("Error: No faces detected in video", file=sys.stderr)
+        sys.exit(1)
+
+    print("\n[4/5] Clustering faces...")
+    clusters = cluster_faces(all_faces, eps=args.eps, min_samples=args.min_samples)
+    print(f"Found {len(clusters)} clusters")
+
+    print("\n[5/5] Generating output...")
+    generate_output(frames, all_faces, clusters, args.output)
+
+    print(f"\nDone! Output saved to: {args.output}")
+    print("  - frames/          : Frames with bounding boxes")
+    for i, cluster in enumerate(clusters):
+        if cluster.id == -1:
+            print("  - unclustered/     : Single-occurrence faces")
+        else:
+            print(f"  - cluster_{i:02d}/     : Face crops for cluster {i}")
+
+    detector.close()
+
+
+if __name__ == "__main__":
+    main()