diff --git a/.github/workflows/draft-pdf.yml b/.github/workflows/draft-pdf.yml
new file mode 100644
index 0000000..971cce1
--- /dev/null
+++ b/.github/workflows/draft-pdf.yml
@@ -0,0 +1,23 @@
+on: [push]
+
+jobs:
+  paper:
+    runs-on: ubuntu-latest
+    name: Paper Draft
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+      - name: Build draft PDF
+        uses: openjournals/openjournals-draft-action@master
+        with:
+          journal: joss
+          # This should be the path to the paper within your repo.
+          paper-path: paper.md
+      - name: Upload
+        uses: actions/upload-artifact@v1
+        with:
+          name: paper
+          # This is the output path where Pandoc will write the compiled
+          # PDF. Note, this should be the same directory as the input
+          # paper.md
+          path: paper.pdf
diff --git a/paper.md b/paper.md
new file mode 100644
index 0000000..fce6408
--- /dev/null
+++ b/paper.md
@@ -0,0 +1,63 @@
+---
+title: 'Dynax: A package for parameter estimation and linearization of dynamical system'
+tags:
+  - Python
+  - control
+  - dynamics
+  - system identification
+  - parameter estimation
+  - linearization
+authors:
+  - name: Franz M. Heuchel
+    orcid: 0000-0002-6084-0170
+    affiliation: 1 # (Multiple affiliations must be quoted)
+  - name: Finn T. Agerkvist
+    affiliation: 2
+affiliations:
+ - name: Department of Electrical and Photonical Engineering, Technical University of Denmark, Denmark
+   index: 1
+date: XX August 2023
+bibliography: paper.bib
+
+---
+
+# Summary
+
+`Dynax` is a Python package for modeling nonlinear dynamical systems,
+identifying their parameters from data, and designing linearizing feedback laws
+for controlling their input-output behaviour.
+
+Features
+
+- Multiple shooting estimation
+- Multiple experiment estimation
+- 
+
+# Statement of need
+
+- nonlinear compensation: no open-source software at all
+
+Currently, there exists no tools in the Python ecosystem that facilitate parameter estimation 
+for differential equation systems (sometimes called "grey-box models"). There exists 
+[`nlgreyest`](https://se.mathworks.com/help/ident/ref/nlgreyest.html) in Matlab's Control 
+System toolbox and [`DiffEqParamEstim.jl`](https://docs.sciml.ai/DiffEqParamEstim/stable/) in Julia.
+
+`Dynax` was designed to be used by researchers, students and engineers. It
+is already used in a course at DTU for estimating the parameters of loudspeaker drivers instead of a commercial black-box solution.
+The combination of design and ease of use make it possible to iteratively test and
+develop dynamical system models for linearization of such transduers.
+
+# Similar software
+
+- Dynamax
+- nlgreyest
+- nlgreyfast
+- DiffEqParamEstim
+
+# Acknowledgements
+
+We acknowledge contributions from Brigitta Sipocz, Syrtis Major, and Semyeong
+Oh, and support from Kathryn Johnston during the genesis of this project.
+
+# References
+