A lot of times when I find myself updating my CV, I end up spending more time struggling with formatting than actually doing the tiny update I wanted. Frustrated by this, I thought that there should be a faster way to manage this process. The solution? Separate the content from the format. That means having a file containing your personal information and another file defining how to present it. This problem is nicely solved by LaTex, a tool widely used in academia for document preparation.

In this article, we’ll combine LaTex with Pandoc to provide a simple way to generate a professional CV from data. This way, the content of your CV is stored in a structured format (YAML), and the design is handled by a LaTex template. Whenever you need to update your CV, you just modify the YAML file and regenerate the PDF.

The workflow

The architecture is simple. Pandoc will read your data from a YAML file and use a LaTex template to format it into a PDF:

YAML (The Data) + LaTeX (The Template) → Pandoc (The Engine) → PDF (The Result)

Here’s how to set it up step-by-step.

1. Prerequisites

You will need two main tools installed on your system:

  • Pandoc, the “universal document converter”. You can read the install instructions here.
  • TeX Distribution that provides the PDF rendering engine. The recommended options are:
    • Windows: MiKTeX
    • MacOS: MacTeX
    • Linux: TeX Live via your package manager.

Verify your installation by running:

1
2

pandoc --version
pdflatex --version

They should print version information. If not, make sure they are correctly installed and added to your system PATH.

2. Create the YAML file

Create a YAML file named, for example, cv.yaml. This file will store your personal information and CV sections, acting as your “Single Source of Truth”. If you get a new job or change your phone number, this is the only file you need to edit.

You need to follow the YAML syntax, but don’t worry, it is designed to be human-readable and easy to write. For reference, you can check the YAML specification.

Here’s a YAML example with my details:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

name: "Xicu Marí Prats"
profile: >
  Research Engineer with 3+ years of hands-on experience in digital hardware
  design. Seeking opportunities in R&D environments to drive innovation.

contact:
  email: "hi@xicu.net"
  website: "xicu.net"
  linkedin: "linkedin.com/in/xicu"

key_skills: [
  "VHDL", "SystemVerilog", "Xilinx Vivado", "Altera Quartus II", "FPGA Design",
  "OpenPiton", "C/C++", "Python", "Cadence", "KiCad", "Altium", "Linux", "Git"
]

professional_experience:
  - position: "Research Engineer"
    company: "HPDSA Group at Barcelona Supercomputing Center"
    dates: "Dec 2024 -- Present"
    details:
      - "Implemented a state-of-the-art DAE accelerator for AI workloads."
      - "Integrated hardware modules into the OpenPiton framework."

education:
  - degree: "Master's Degree in Electronic Engineering"
    institution: "Universitat Politècnica de Catalunya"
    dates: "Sep 2022 -- Sep 2025"
    details:
      - "Overall grade: 9.1 / 10"

Note: If your data contains special LaTeX characters like & or %, remember to escape them (e.g., R&D) to avoid build errors.

3. Create the template

Now, let’s define the style of your CV. Pandoc treats this as a template, where variables from the YAML will be injected.

A minimal example could look like the following. Here, the $variables$ will be replaced by the contents of your YAML file. Loops like $for(experience)$ … $endfor$ let you repeat blocks for each item in a list. Save this as cv.tex:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

\documentclass[11pt,a4paper]{article}
\usepackage{geometry}
\geometry{margin=2cm}
\usepackage{enumitem}
\usepackage{hyperref}

\begin{document}

\begin{center}
    {\Huge \textbf{$name$}} \\[0.5em]
    \href{mailto:$contact.email$}{$contact.email$} | \href{https://$contact.website$}{$contact.website$} | \href{https://$contact.linkedin$}{$contact.linkedin$}
\end{center}

\vspace{1em}

\section*{Education}
$for(education)$
\textbf{$education.degree$}, $education.institution$ \hfill $education.dates$ \\
$endfor$

\section*{Experience}
$for(professional_experience)$
\textbf{$professional_experience.position$} --- $professional_experience.company$ \hfill $professional_experience.dates$ \\
\begin{itemize}[leftmargin=*]
$for(professional_experience.details)$
    \item $professional_experience.details$
$endfor$
\end{itemize}
$endfor$

\end{document}

4. Generate the PDF

With everything set up, you can generate your CV PDF with:

1
2
3
4

pandoc cv.yaml \
    --template=cv.tex \
    -o cv.pdf \
    --pdf-engine=pdflatex

This command tells Pandoc to:

  1. Load the data from cv.yaml.
  2. Use cv.tex as the template.
  3. Produce a PDF with pdflatex.

5. The full tool

Now that you know how this solution works under the hood, it’s time to introduce cv-cli, a command-line tool available in the following public repository:

👉 https://github.com/XicuM/cv-cli

To use this utility, you must first install it. Download the source code from the repository and open a terminal from the main code directory. Once there, run:

1

pip install --user -e .

Once installed, you can start using it. The main commands are:

1
2
3
4
5
6
7

cv init               # Initialize project with templates to fill in
cv cv -l en           # Create a CV in English
cv letter acme -l en  # Create a letter in English for the company "acme"
cv template cv        # Edit the CV template
cv build              # Build the entire project
cv build -t cv -l en  # Build only the CV in English
cv clean              # Clean the project cache

If you try cv init and then cv build, an output folder will be generated with the following documents:

cv-en.pdftemplate-en.pdf

With that, you’ll have a professional CV in hand! From here, you just need to tweak the LaTeX file to your preferred style once and update the .yaml file whenever you finish a professional stage.

Happy crafting! If you run into any issues with Pandoc or LaTeX environments, feel free to drop a comment below.