Comprehensively mapping the relationship between genotype and phenotype offers essential insights into how a cell’s state arises from its genetic components. Toward this goal, we generated an expressed genome-scale CRISPRi perturbation cell atlas in KOLF2.1J human induced pluripotent stem cells, mapping transcriptional phenotypes associated with 11,692 perturbed genes across >2.5 million single cells. Using correlations among perturbed phenotypes, we created a cell map of the pluripotent state, demonstrating rich recapitulation of functionally related protein complexes. We then explored the atlas to uncover metabolic factor ZBTB41 and pluripotency regulator RNF7, validating their functions through metabolic tracing, immunofluorescence and protein–protein interaction assays. Lastly, we leveraged the atlas to generate a genome-scale screen of A-to-I RNA-editing modulators assayed through direct transcriptome-wide RNA editing, uncovering and mechanistically validating DBR1 as a potent regulator. Taken together, our data provide a comprehensive resource for interrogating the regulatory networks governing pluripotency, which is accessible at https://y-doctor.github.io/KOLF2.1J_Perturbation_Cell_Atlas/.