Description
I am looking for an experienced AI engineer to fix one critical issue in an existing AI agent. The AI agent is already built and working correctly in most workflow sections. It is a document processing and structured data extraction agent. The agent receives uploaded documents, analyzes the content, extracts required fields, validates the extracted information, and generates a structured final output for the application. However, there is one critical issue in the validation and final-output generation section of the workflow. In some cases, the agent marks the process as completed and returns a final result even when required extracted fields are missing or the validation step has failed. This causes the application to treat incomplete or invalid document data as successfully processed. This is not a full rebuild. I only need an experienced engineer to investigate the existing workflow, identify the root cause, and fix this specific issue. Detailed technical documentation is attached. Please review it before applying. Timeline: Urgent / start as soon as possible Please apply only if you have experience AI agents, LLM workflows, tool/function calling, validation logic, and state management.