Hey Crunchers,
We’re cruising through July, but it’s anything but quiet here. From 3,000+ model submissions in the structural break challenge to lab updates from the Broad Institute, the Crunch machine is in full swing. Jean’s been making the rounds and our Crunchers are turning heads on every leaderboard.
Let’s get into it 
ADIA Lab Structural Breaks Challenge: 3,000+ submissions and counting
Over 45 teams and hundreds of individuals are now competing in the ADIA Lab Structural Breaks Challenge, with more than 3,000 model submissions already in!
If you haven’t jumped in yet, now’s the time.
Need help getting started? Watch this walkthrough by our CEO, Jean Herelle →
Watch the Submission Tutorial
Join the challenge → Enter the Competition
P.S. We’ve made a small adjustment to the Structural Breaks competition timeline to reflect evolving plans. Originally, we shared exact dates for final scoring and the winner announcement. Those have now been replaced with more flexible timelines:
➛ Final evaluation will take place toward the end of October
➛ Winners will be announced during the ADIA Lab 2025 Symposium, which runs from October 27 to 29
Autoimmune Disease Machine Learning Challenge: Timeline & Sharing Policy
We have an official update from the Eric and Wendy Schmidt Center at the Broad Institute of MIT and Harvard on what’s next in the Autoimmune Disease ML Challenge.
Timeline:
-
Phase 1: Model Review (1–2 months)
Submissions are being analyzed across all hold-out tissue sections.
ETA: August 2025 -
Phase 2: Experimental Validation (3–4 months)
Selected models will undergo wet-lab validation.
ETA: Early 2026 -
Final Results:
You’ll receive official rankings and lab findings once the experimental phase concludes.
Sharing Guidelines:
You can share your modeling approach and leaderboard placement.
You can’t share data, extra plots, private metrics, or performance beyond what’s public.
Payout reminder:
All payouts will be distributed at once when the broad-3 results are announced.
DataCrunch Equity Market Neutral: Leaderboard
Don’t sleep on the DataCrunch Equity Market Neutral Challenge.
Crunchers continue to battle it out in one of our most competitive arenas yet—designing market-neutral models that consistently generate alpha across regimes. With multiple leaderboard shifts this week, the top ranks are tighter than ever.
If you’re looking for a challenge that’s as real-world as it gets, this one’s for you. Live capital deployment. Real returns. Zero room for noise.
Check the leaderboard → DataCrunch Leaderboard
Cruncher Spotlight: Team Cellmates
In this week’s Autoimmune Challenge II spotlight, we’re featuring Team Cellmates.
Their hybrid approach blends contrastive learning, nearest-neighbor search, and deep domain intuition.
Team Members:
-
Marios Gavrielatos: Bioinformatics researcher, long-read sequencing wizard
-
Konstantinos Kyriakidis: Pharmacist turned bioinformatics engineer
Approach Highlights:
-
Used their Crunch1 model to predict 460 gene expressions from H&E images
-
Applied FAISS to find 5 closest single-cell profiles per spatial sample
-
Created a (5, 2458) matrix per sample combining predicted + real gene values
-
Fed that into a downstream model to infer target expression
An elegant, multi-stage solution that’s equal parts vision and execution.
Huge kudos to Team Cellmates!
Crunch Team Spotlight: Grégoire Joins as Chief of Staff
We’re thrilled to welcome Grégoire Colcombet to Crunch as our new Chief of Staff!
Grégoire previously worked at McKinsey and later led ops at Zealy. At Crunch, he’ll drive strategic initiatives and help scale the organization with clarity and precision.
“I’m really excited to join such a sharp and passionate team. The business model is refreshingly original, and I’m convinced this platform can fundamentally change how machine learning and financial modeling gets done.”
Let’s go, Grégoire 
Jean Herelle: On the Mic
SpineDAO @ Roundtable with Mario Nawfal
Jean joined Roundtable Space to discuss how collective intelligence can tackle spine care and unlock new frontiers in decentralized health.
Coordination in Science @ Bonfires.ai
Jean spoke with Bonfires AI on how Crunch enables structured collaboration in science through on-chain modeling.
Founders Roundtable #50 @ YAY Network
A conversation on how AI is driving fundamental shifts in crypto and institutions.
Have Questions?
We’re prepping for the next Community Call and we want to hear from you!
Submit your questions or upvote the ones you care about the most:
—
That’s all for now. Stay curious. Stay Crunchin’.
— Team CrunchDAO