This blog post presents a novel agentic workflow designed to enhance observational causal inference (OCI) by leveraging software agents. It emphasizes the role of human oversight in ensuring the validity of analytic results while maintaining rigorous methodological standards. The authors introduce an open-source tool to model OCI workflows, share evaluations demonstrating its effectiveness, and discuss the importance of transparency and diagnostics in causal analysis. Additionally, the post includes a case study on estimating the impact of new entertainment types at Netflix, illustrating the application of their methodology in practice.