An end-to-end ecosystem that enables the secure sharing of proprietary datasets alongside the high-performance compute and orchestration layer to train models.
Bellevue, United States
Joined September 2026
- Tweets8
- Following36
- Followers3
- Likes4
The biggest bottleneck in AI for biology isn’t the code. It’s the data deadlock.
AI developers need exclusive, high-value experimental datasets. Data owners have them but lack a secure way to share. Innovation stalls.
We’ve built the way out: harelldata.com 🧵
#ai #biotech
1
8
A few of the datasets available on Harell Cloud:
• HLA-A*02:01-restricted TCR–pMHC functional specificity (>1,000,000 paired α/β TCRs)
• HLA-B*27:05-restricted TCR–pMHC (>27,000 paired α/β TCRs)
• Dual-specific scFv SSM × LILRB4, TIGIT, TNFRSF9/4-1BB
• ...and more
1
4
You train inside the platform, so the data stays secure and the model stays yours.
Want to learn more about available datasets or how the platform works?
Let’s get in touch: harelldata.com/contact-us/
2
New - CASP-DREAM TAP Challenge - TCR-pMHC Activation Prediction (TAP)
synapse.org/Synapse:syn77965…
@AdaptiveBiotech @DR_E_A_M @Sagebio
3