The KiraLOGIC Platform
KiraLOGIC is our AI-powered Large Phenotype Model (LPM), built to design multiplex edits that keep immune cells functional under TME pressure.
It learns from real phenotypes—cytokine release, spheroid killing, post-stimulation persistence—so our edits are based on how cells actually behave, not just what they express.
Turning Data Into Durable Immunity
KiraLOGIC turns real biology into smarter edits using a closed-loop system powered by high-dimensional assays and machine learning.
‘Intervention’ includes gene edit combinations applied to cells.
Outputs shown are representative, not exhaustive
How It Works
1
Biology-Informed Inputs
Trained on organoids, cytokines, scRNA-seq, and phenotypic cues—KiraLOGIC starts with the right biology to ask the right questions.
2
Multiplex Edit Prediction
Screens billions of edits to identify synergistic, low-toxicity combinations that resist TME-driven suppression.
3
Wet Lab Validation & Iteration
Top edits are tested in vitro and in vivo using high-content assays to sharpen predictions and speed translation.
4
Improves with Every Dataset
KiraLOGIC is a closed-loop platform—constantly learning from new biology to refine edits and expand therapeutic reach.
Why It Matters
Built to Persist
KiraLOGIC doesn’t just optimize edits—it designs programmable durability, embedding resistance into the therapy rather than patching it on. This becomes a logic layer for immune cells that endure where others fail.
Scalable by Design
Each program fuels the KiraLOGIC flywheel—refining smarter edits across αβ T, NK, and γδ T cells. Indication-specific insights become transferable resistance strategies, expanding reach without added complexity.
The Engine Behind Programmable Immunity
We’re not just predicting biology.
We’re designing therapeutic function.
​
KiraLOGIC is the backbone of our programmable cell therapy platform—
where suppression resistance is engineered, not hoped for.