Jason Givati

About

I work on the failure mode most monitoring misses: the run that completes, reports success, and produced nothing. No exception is thrown, so nothing in the error handling fires. I audit for it by reading execution history from outside the instance, which catches what reviewing the workflow cannot — on 2.36.7 a node on a branch whose condition stops matching is absent from the run data entirely while the execution still returns success. The audit is a fixed fee: $500, one week of fixed scope, read-only, and the written report is yours either way.

n8n experience