Tesseract Hackathon

Dec 6, 2025 – Jan 5, 2026

Building differentiable simulation pipelines for the next generation of scientific computing

Your Task#

Build something cool with Tesseract, an open-source software ecosystem for creating end-to-end differentiable pipelines that bridge research and simulation workflows. Win awesome prizes, or at least some swag.

Tesseract lets you compose computational components while maintaining gradient flow through complex pipelines. This hackathon is your chance to get hands on with Simulation Intelligence—a unified framework where differentiable programming, simulation, optimization, and ML converge.

We're looking for creative applications that push boundaries within inverse problems, multi-scale modeling, automated experimental design, or anything we haven't imagined yet.

Recognition & Prizes#

First Place

$1,000

Plus a research collab opportunity with our team

Second Place

$500

Plus feature spot on our technical blog

Social Media Bonus

$100

Team whose posts get the most reactions

Honorable Mentions: Outstanding submissions become case studies in our docs and get community recognition.

🎉 Everyone Wins! 🎉

All eligible submissions receive exclusive Tesseract Hackathon swag, including a limited-edition T-shirt.

Subject to availability and shipping constraints, see Terms & Conditions.

Timeline#

Dec 6
Hackathon Begins
Dec 15
Virtual Office Hours
(Optional Q&A Session)
Jan 5
Submissions Deadline
11:59 PM Anywhere on Earth
Jan 19
Winners Announced

How to Participate#

Eligibility#

Open to people of all backgrounds including hobbyists, PhD students, postdocs, research engineers, and industry researchers. Solo or teams up to 4 people.

What to Submit#

  • A project that meaningfully uses Tesseract's differentiable pipeline capabilities. Check out Ideas to Explore for inspiration and datasets to get started.
  • Code written during the hackathon period, licensed under Apache 2.0.
  • Public GitHub repo with clear documentation—README with setup instructions, architecture overview, and usage examples.
  • Short technical writeup (2-4 pages) or detailed README explaining your approach.
  • Optional: Demo video (5 min max) showing what you built and why it matters.

What We're Looking For#

Projects will be judged on:

  • Scientific Impact (35%): Does this solve a real problem? Could it enable new research?
  • Technical Execution (30%): Smart use of differentiable programming and clean pipeline design.
  • Reproducibility (20%): Can others run and build on your work?
  • Communication (15%): Clear explanation of what you built and why.

How to Submit#

Post on LinkedIn and X/Twitter with your GitHub link. Tag us @ Pasteur Labs & ISI / SimAI4Science, and use the hashtag #TesseractHackathon. That's it!

Registration#

To participate in the Tesseract Hackathon, please complete our registration form. Registration is free and helps us keep you updated with important announcements, details on office hours, and resources throughout the hackathon.

Register Now

By registering, you agree to our Terms and Conditions. Registration is required to be eligible for prizes.

Resources#

Everything you need to start building with Tesseract:

Ideas to Explore#

Some application areas with datasets and resources to get you started:

  • Inverse Design for Structural Engineering: Optimize jet engine brackets, structural components, or metamaterials using gradient-based topology optimization. Smart Design Lab datasets provide real engineering problems with complex constraints.
  • Differentiable Fluid Dynamics: Build surrogate models for computational fluid dynamics (CFD) or solve inverse problems in fluid mechanics. Try aerodynamic shape optimization using SU2 or OpenFOAM datasets.
  • Material Discovery & Molecular Design: Use differentiable molecular dynamics to discover new materials or optimize molecular properties. Explore Materials Project datasets or RDKit for molecular structures.
  • Climate & Weather Modeling: Build differentiable weather prediction models or optimize climate intervention strategies. Use ECMWF datasets or NOAA climate data.
  • Robotics & Control: Design differentiable controllers for robotic systems or optimize trajectories through complex environments. Try OpenAI Gym or PyBullet physics simulations.
  • Biomedical Applications: Inverse problems in medical imaging (CT, MRI reconstruction), drug response prediction, or biological system modeling. Explore Cancer Imaging Archive or ChEMBL drug datasets.
  • Energy Systems Optimization: Optimize power grid configurations, battery charging strategies, or renewable energy systems. Use NREL grid data or OpenEI energy datasets.
  • Neural Architecture Search: Use differentiable techniques to automatically design neural network architectures optimized for specific scientific computing tasks.
  • Multi-Physics Coupling: Build end-to-end differentiable pipelines that couple multiple physics domains (thermal-structural, fluid-structure interaction, electro-mechanical systems).

... and anything else we didn't anticipate! We love creative submissions that push boundaries in unexpected directions. If you have an idea that combines differentiable programming with your domain expertise, go for it!

Get Started#

Jump straight in with our hackathon starter template. We've set up a repository with pre-configured Tesseracts, example scripts, and common workflow patterns to help you hit the ground running.

📦 GitHub Template Repository 📦

Quickstart

Clone the template and get started in minutes:

# Clone the hackathon template
git clone https://github.com/pasteurlabs/tesseract-hackathon-template.git
cd tesseract-hackathon-template

# Set up your environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the example script
python main.py

New to Tesseract? Check out the Resources section for tutorials, documentation, and research papers to help you understand the framework.

Join the Challenge#