Generative protein design has crossed a threshold. The best AI methods now produce functional candidates at rates that were unimaginable two years ago.
Which means design is no longer the bottleneck. The loop is. Every step has to produce signal clean enough for the next cycle to learn from. Design. Expression. Characterization. Analysis. When any link breaks, models get confident on noise, the next round is worse than the last, and the whole stack quietly degrades.
The teams shipping drugs in 2026 aren't winning on model size. They're winning on loop discipline.
Tight coupling between design and bench. Fast, clean characterization. Data the next training run can actually use. That's the work.
On June 5, four of the teams who've figured this out come together for one focused afternoon. No hype. No vendor pitches. Just the people doing the work, talking about what closes, and what doesn't.
FREE TO ATTEND · LIMITED CAPACITY