Spatial assays are becoming the core driver of fabled compounding data generation curves in biology, allowing scientists to look at the molecules in their tissue and explore how this state pairs with visual features to understand disease, development and so forth. Unlike most

Agents have gotten quite good at math and programming, but require intentional engineering to translate well to biological data. Latch spatial agents are tailored to the details of each kit, in direct collaboration with the companies that sell them, so they have the specific

So what do agentic workflows look like in practice? We’ve compiled very detailed flows for four different assay types in the blog post linked at the end. A tailored agent is built for each kit or machine type. Scientists begin with raw data, send instructions or questions to a

2/ Run differential chromatin analysis on external machines

3/ Pull out genomic tracks and embed IGV in the browser

4/ Cluster and annotate cell types

Encourage all to dig into the materials we put together here. Prefer to show not tell: https://blog.latch.bio/p/agent... While there is some prior art, believe our approach here is somewhat unique. Surgical products with guardrails, that scientists use to do very precise (but difficult)

@kenbwork The speed of innovation is incredible ! A specific area that I am weary about and that I see many models struggle with is hallucinations, what your approach to reduce/eliminate these occurences?