I’ll review one existing AI-assisted workflow and identify where it is likely to fail, drift, overreach, or create unnecessary manual work.

The review focuses on the process around the model—not hype about the model itself. I look at inputs, decision points, prompts/instructions, human approval boundaries, handoffs, error handling, source-of-truth problems, and completion criteria.

a map of the current workflow

identified failure and drift points

missing or weak human review points

unclear instructions, state transitions, or decision boundaries

source-of-truth / context risks

prioritized recommendations

a revised workflow outline showing stronger controls

a description of the workflow

relevant prompts or instructions

screenshots, documents, or examples of outputs

what the workflow is supposed to accomplish

any known problems or failure cases

This service covers one bounded AI-assisted workflow. It does not include building custom software, agents, integrations, or automations.