diff --git a/README.md b/README.md
index d4b3baf..0b486c0 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,17 @@
-# Setup
+# LLM-NVR
+
+This project is a proof of concept for using multi-modal LLMs to integrate with NVRs and perform narration, explanation on top of object detection and tracking.
+
+## Setup
 
 1. Install python3 and activate the venv: `python3 -m venv venv`
 2. Install requirements: `pip install -r requirements.txt`
-3. Run docker compose to bring the frontend and backend up: `docker-compose up -d`
-4. Run the jupyter notebook in analyze.ipynb
\ No newline at end of file
+3. Edit configs in frigate/ subdirectory and bring up Frigate with docker compose: `cd frigate && docker-compose up -d frigate nvr`
+4. Run docker compose to bring the frontend and backend up: `docker-compose up -d`
+5. Run the jupyter notebook in analyze.ipynb
+
+## Demo
+[Link](https://x.com/arcyleung/status/1835194959290327250)
+
+## Architecture
+<img src="arch.jpg" alt="architecture" width="80%" style="margin: auto"/>
\ No newline at end of file
diff --git a/arch.jpg b/arch.jpg
new file mode 100644
index 0000000..bdd2b41
Binary files /dev/null and b/arch.jpg differ