This is a little experiment. I wanted to know more about where my CVs go and what happens to them. Since they contain quite some links to my website, linkedIn, Gihtub etc... I thought, let´s try to apply e-mail tracking strategies to PDFs.

• A PDF generator that replaces all the links with trackable links
• A redirect service that redirects the user from the tracking links to the actual links with utm_codes.
• A service that tries to find the Company name by their IP address

• AWS account
• AWS CLI
• Serverless Framework
• Python 3.x
• Node.js

npm run setup

This will install the python virtual env & dependencies. JavaScript Dependencies for Serverless Framework

Rename .env.sample to .env & change according to your needs.

STACK_NAME="redirect-service-test"

Name of the AWS CloudFormation Stack. Choose a unique name to make sure it does not conflict with other stacks.

REDIRECT_URLS_TABLE="redirectUrlTable"

This table stores the redirect URLs and the information about the original destination, individual CV PDFetc.

REDIRECT_TRACKING_TABLE="redirectTrackingTable"

This table will hold the actual tracking data of the click events

IP_TABLE="ipTable"

This table is a cache for additional IP data

URL_NOT_FOUND_REDIRECT="https://thomaszipner.com/sorry-the-redirect-failed"

This key is for testing and actually works, but is highly rate limited. Create an account at https://ipdata.co/ to get your own API key.

IPDATA_API_KEY="test"

CV_VERSION="1.0" Just some version number, so you can see what version was clicked.

BASE_PDF="cv-with-links.pdf" FINAL_PDF="You Name - CV.pdf"

The name of the input PDF and the name to your output PDF. You can use a file path, if it isin another folder.

This is the url that will replace your links from the PDF. If you do not want a custom domain, the just use the ApiGateway output from serverless after running "sls deploy"

REDIRECT_SERVICE_BASE_URL="https://rizyzb6wfb.execute-api.eu-central-1.amazonaws.com/prod-redirect-service/"

You can use a custom domain, but theres is some more config to do.

1. Uncomment this part in serverless.yml

custom:
  customDomain:
    # activate this if you want to use a custom domain
    # domainName: ${env:REDIRECT_SERVICE_DOMAIN}
    # basePath: ''
    # stage: ${self:provider.stage}
    # createRoute53Record: true

Your TLD needs to be registered in Route 53.

Once you have your domain, request a new certificate with the AWS Certificate Manager. Note that you'll need to be in region us-east-1. This is the only region that works with API Gateway.

Then run:

sls create_domain

Now you can set

REDIRECT_SERVICE_DOMAIN="rdr.yourdomain.com"
REDIRECT_SERVICE_BASE_URL="https://rdr.yourdomain.com/"

Run:

source venv/bin/activate

to activate the Virtual Environtme, so the Servlerless Framework knows all the Python dependencies and then:

sls deploy

Lean back and watch the magic! It may take a few minutes.

Now take the first output under "endoints" and insert this as REDIRECT_SERVICE_BASE_URL in your .env file.

If you now visit that link, you should see your redirect service in action. Since we have no links yet, you are redirected to URL_NOT_FOUND_REDIRECT.

Go the pdf-generator folder.

cd pdf-generator

There already is a test PDF that you can use, but you can replace it with your own. If the name or path is different, just change BASE_PDF in your .env.

To generate a new tracked PDF and the redirect links type:

python main.py "{SOME-IDENTIFIER}"

Replace {SOME-IDENTIFIER} with whatever makes sense to you. I usually use e.g. the name of the company or recuiter I send this CV. Besides the name in your DynamoDB, it genereates utm_codes with the utm_campaign set to a hash of {SOME-IDENTIFIER}, so if you use Google Analytics, you can see this as your campaign name. 😃

Now open your PDF (You Name - CV.pdf is in the same folder or where you specified it to be) and click some links.

Your tracking data lives in the DynamoDB tables. You could write an API against it to feed some frontend to make it look nice. Mayba I will do that when I have some time.

Holds the info about the urls for the redirecting service.

Holds the inofrmation about each click on a link, so you can see from where it came. Additional company data gets added by the whois-service if it finds any. This is pointless if it is an ISP, but can be very interesting if it the company you just apllied to.

is basically a cache for the IP lookups, so we need less API calls.