diff --git a/README.md b/README.md
index 71addeea..f97f6bf1 100644
--- a/README.md
+++ b/README.md
@@ -1,15 +1,18 @@
-### 🦜 PARAT: CSET's Private-sector AI-Related Activity Tracker
+# 🦜 PARAT: CSET's Private-sector AI-Related Activity Tracker
 
 This repository contains the code that generates [PARAT](https://parat.cset.tech).
 
-[company_linkage](/company_linkage) contains the code that generates the data behind PARAT: 
+* [company_linkage](/company_linkage) contains the code that generates the data behind PARAT:
 company resolution across datasets, publication counts, etc.
 
-[web](/web) contains the code that generates the PARAT website from the company_linkage data.
+* [web](/web) contains the code that generates the PARAT website from the company_linkage data.
+
+
+## Setup
 
 The project python dependencies are listed in `requirements.txt`. Before running code in this repo:
 
-1.) Make a new virtualenv:
+1. Make a new virtualenv:
 
 ```bash
 python3 -m venv venv
@@ -17,7 +20,10 @@ source venv/bin/activate
 pip install -r requirements.txt
 ```
 
-2.) `export GOOGLE_APPLICATION_CREDENTIALS=<path to your credentials>` - a service account json. You should have at 
+2. `export GOOGLE_APPLICATION_CREDENTIALS=<path to your credentials>` - a service account json. You should have at``
 least BQ reader permissions, if you are a CSET user. If you are not a CSET user, then you will not be able to
 run some of this code as-is, since it depends on our internal BigQuery datasets. Please contact us if you are
-interested in collaborating or spot any issues.
\ No newline at end of file
+interested in collaborating or spot any issues.
+
+3. [Install and run the web interface](/web/README.md#web-interface)
+
diff --git a/web/README.md b/web/README.md
index 31c3f432..0d67b697 100644
--- a/web/README.md
+++ b/web/README.md
@@ -1,30 +1,41 @@
-### Web artifacts and data preprocessing code for PARAT
+# Web artifacts and data preprocessing code for PARAT
 
 This directory contains the code that processes the data generated in [company_linkage](../company_linkage) for use in the PARAT website.
 It also contains a Gatsby project that contains the PARAT website generation code.
 
+## Data updates
+
 To update the text that is used in the various explanatory tabs, edit the markdown files in `raw_data/text/`.
 To update the tooltips, edit `src/static_data/tooltips.js`.
 
 To populate the necessary data from a raw clone of this repository, run:
 
-1.) `python3 scripts/mk_tab_text.py`. This will read the markdown text files in `raw_data/text` and output them
+1. `python3 scripts/mk_tab_text.py`. This will read the markdown text files in `raw_data/text` and output them
 as a javascript object containing html snippets in `src/static_data/text.js`.
 
-2.) Grab the raw data and reformat it into a javascript object. You will need a service account with translation 
+2. Grab the raw data and reformat it into a javascript object. You will need a service account with translation
 and BigQuery reader permissions. To fully regenerate everything including images and Google Finance links 
 (which take ~1.5 hours to generate), run: 
 `python3 scripts/retrieve_data.py --refresh_raw --refresh_images --refresh_market_links`. Run
 `python3 scripts/retrieve_data.py -h` for more detail on what these parameters do.
 
-Next, to run the site in development mode, do:
+
+## Web interface
+
+The new (v2) interface for PARAT is in the `gui-v2/` directory.
+
+### Development server
+To start the development server:
 
 ```bash
-cd parat 
-gatsby develop
+cd gui-v2/
+npm install
+npm run develop
 ```
 
-You should be able to view the site at `localhost:8000`.
+The PARAT v2 interface will be available at `localhost:8550`.
+
+### Deploying release
 
 When any changes are ready for deployment, do: