Resume

This repository contains my most up-to-date resumes. The PDFs themselves are in the pdf/ directory, while the source needed to compile them is in the root directory. Because of the complexities of compiling multiple PDFs, the code is based on a custom templating engine I wrote, called progres. To compile this code, you will need to have that installed.

Compiling

To generate the PDFs, run

progres -o pdf

You can also use

make

These will generate all the PDFs simultaneously and move them to the pdf/ directory.

How does this work?

The resume that led you here is written using a custom resume templating system I wrote, called progres. The readme for that project (linked above) goes into detail, but in brief:

• data.json - Contains all the raw data. Different PDFs use different parts of this data based on the role. Items have a tags attribute: these are used for filtering.
• configs.json - This file describes, for each role, what tags are/aren't set. These are simply arrays of Python code, and they set different variables (which are the same as the tags) for each config. Items are only shown in a config if all the tags are set to True.
• computeRole.py - This adds additional rules; the main one being that for the "master" resume, all tags are enabled.
• preamble.tex - This is the preamble section of the intermediate LaTeX document, see below.
• spec.py - Describes how to generate the LaTeX files for each config.
• Specfile - A set of commands that describes how to build the resumes. It first imports a few necessary files described above. Each PARSE command is a call to one of the parse_ functions from spec.py, and any arguments are forwarded as-is. Lines that don't start with a recognized command are treated as Python code.

With these files in place, the "main" file is the Specfile. progres performs the following conversions:

Specfile --> Python files --> LaTeX files --> PDFs

progres is an interpreter that parses the commands and Python code from Specfile, and generates one Python file for each config. These are then run in parallel, and each of them generates a LaTeX file. These are compiled in parallel (generating the PDFs), and the intermediate files are deleted.

FAQ

Why not a single resume?

To generate tailored resumes for different roles. When applying to ML roles, for example, a lot of the research-related data is pulled in; these are excluded for SDE.

This is too complex/over-engineered!

Correct, but it's cool. I have crazy ideas; sometimes they work. This is one of those times.

I had a very specific need and wanted an easy way to generate up-to-date versions of each PDF.

Why not use [alternative]?

It wasn't quite as flexible as I wanted, most likely.