Crunch Weekly Recap
Hey Crunchers,
In an age where algorithms shape markets, science, and society, one question keeps coming back: Can intelligence be decentralized and still move forward?
This month we are one step closer to answering it!
Across oceans and time zones, crunchers built models on publicly shared clinical datasets from the Broad Institute over the last 6 months. Now bear in mind that they involved spatial transcriptomics, histology images, and high-dimensional gene expression. The kind of data that’s usually kept behind institutional gates.
Now, these models are entering lab validation.
What this proves is simple but profound: you don’t need to be inside a HQ to shape science. If coordination is strong and incentives are aligned, the world’s sharpest minds can deliver breakthroughs that were once possible within centralized labs or quant desks.
Crunch is that coordination layer. What collective intelligence actually looks like.
Now, let’s dive into this week’s updates 
Have Questions? We’re Listening
First things first, we hear you. We see some community members dropping questions about future plans, roadmap and more and we want to have them answered. We would like to aggregate all these questions from you to answer them over an AMA.
Just drop them here
ADIA Lab Structural Break Challenge Is Live
Have you noticed that the new ADIA Lab Crunch is now live? A $100K time-series competition designed with experts from ADIA Lab. The task is to detect structural breaks before they blindside your models. Submissions are open through November 15th.
Not sure if you know that you can literally submit your first model in 3 minutes and enter the leaderboard using the kickstarter notebook.
Autoimmune Challenge → Final Selection Begins
Your models are now entering a real-world clinical workflow at the Broad Institute of MIT & Harvard, one of the most prestigious cancer research centers globally.
Here’s the updated timeline shared by the team at Broad:
-
2–3 months: Final result analysis + gene selection
-
3–4 months after that: Gene panels ordered + wet-lab validation begins
-
Total: 5–7 months from now until the last milestone
Note: Payouts will be distributed at the end, after final results come in.
DataCrunch Leaderboard Update
Another week, another leaderboard. See who’s topping the charts in our weekly predictive competition.
Payouts Are Rolling Out!
We’re closing out spring with a fresh round of rewards for top Crunchers across multiple challenges:
- DataCrunch April: View Payouts
- DataCrunch May: View Payouts
- Mid+One Round 2: View Payouts
Plus, the ADIA Lab Causal Discovery rewards are ready to be sent to all the winners of the challenge! Congrats to all who crunched and earned!
Peter Cotton at UCSB
At UCSB’s Distinguished Lecture, our CSO Peter Cotton explored what happens when humans lose control of data science.
Podcast: ML, AI & Collective Intelligence
Peter also joined MLBoost to talk about agent alignment, crowd-led science, and the future of AI markets.
Philipp Wassibaeur at Solana Speed NYC
Our CTO repped Crunch at Solana Speed, alongside MetaMask, Arcium, and more. Talking truth, crypto, and compute. Recaps coming soon.
Jean at ADIA Lab Summer School, Granada
From ETH Zurich to Spain, Jean Herelle, CEO of Crunch has been busy bringing structural breaks to the research forefront!
Video drop soon.
Weekends Are for Crunching!
There’s something about weekends. Uninterrupted time, fewer tabs, and just enough quiet to focus on what really matters. That’s when the best ideas emerge. That’s when great models get built. If you’re staring down a notebook or a time series this weekend, you’re not alone fren!
We’re Hiring
Join a team that’s rewriting the rules of predictive modeling.
Here are the open roles:
• Quant Dev
• Business Development Manager
• Developer Relations Lead
• Executive Assistant
• Product Manager
Meme of the Week
For the culture!
That’s all for this week.
Questions incoming, models evolving, and more alpha just ahead. Let the crunch continue.
— Team Crunch