Step 2: Connect Google Docs

The evaluation agent uses two Google Docs tools to fetch the question paper and answer key.

1. Open Fetch Question Document and set Operation to get.
2. Set Document URL to =https://docs.google.com/document/d/[YOUR_ID]/edit.
3. Credential Required: Connect your googleDocsOAuth2Api credentials in Fetch Question Document.
4. Open Retrieve Answer Key and set Operation to get.
5. Set Document URL to https://docs.google.com/document/d/[YOUR_ID]/edit.
6. Credential Required: Connect your googleDocsOAuth2Api credentials in Retrieve Answer Key.

Note: Fetch Question Document and Retrieve Answer Key are AI tools used by Evaluation Orchestrator. Credentials are configured on these tool nodes, while the agent references them during execution.

Step 3: Set Up AI Processing

These nodes analyze the uploaded answer sheet, orchestrate evaluation, and parse structured output.

1. Open Image Response Analyzer and set Operation to analyze, Resource to image, and Input Type to binary.
2. Set Binary Property Name to Upload_Answer_Sheet and Text to =The image is an answer paper of a student, you need to analyze and pick each and every answer along with section name, question number and student name..
3. Credential Required: Connect your googlePalmApi credentials in Image Response Analyzer.
4. Open Gemini Chat Engine and set Model Name to models/gemini-2.5-pro.
5. Credential Required: Connect your googlePalmApi credentials in Gemini Chat Engine.
6. Open Evaluation Orchestrator and set Text to =Marks of the student:\n{{ $json.content.parts[0].text }}\nExaminer Name:{{ $('Exam Submission Form').item.json['Examiner Name'] }}\n\n.
7. Ensure Prompt Type is define and Has Output Parser is enabled.
8. Open Structured Result Parser and keep the JSON Schema Example aligned with the required output format shown in the node.

Note: Gemini Chat Engine is connected as the language model for Evaluation Orchestrator — ensure credentials are added to Gemini Chat Engine, not the agent. Structured Result Parser is a sub-node used by Evaluation Orchestrator and does not take credentials directly.

Step 4: Configure Output/Action Nodes

Results are appended to summary and detailed Google Sheets, with a code step reshaping data for per-question rows.

1. Open Append Summary Row and set Operation to append.
2. Select the target spreadsheet and sheet: Document ID [YOUR_ID] and Sheet Name gid=0 (cached name Sheet1).
3. Map columns to expressions, for example Total Marks to {{ $json.output['Total Marks'] }} and Student Name to {{ $json.output['Student Name'] }}.
4. Credential Required: Connect your googleSheetsOAuth2Api credentials in Append Summary Row.
5. Open Merge Results to JSON and keep the JS Code as provided to transform the structured output into per-question rows.
6. Open Append Detailed Scores and set Operation to append.
7. Select the detailed sheet: Document ID [YOUR_ID] and Sheet Name 760946435 (cached name Scorecard).
8. Map fields such as Status to {{ $json.Status }} and Question Number to {{ $json.Question }}.
9. Credential Required: Connect your googleSheetsOAuth2Api credentials in Append Detailed Scores.

Execution Note: Evaluation Orchestrator outputs to both Append Summary Row and Merge Results to JSON in parallel. Merge Results to JSON then feeds Append Detailed Scores.

Step 5: Test and Activate Your Workflow

Run a full submission to confirm AI extraction and sheet updates work as expected.

1. Click Execute Workflow and submit the Exam Submission Form with a sample answer sheet image.
2. Verify Image Response Analyzer produces extracted text from the uploaded image.
3. Confirm Append Summary Row adds a summary entry and Append Detailed Scores adds per-question rows to your sheets.
4. If output is missing, check that Fetch Question Document and Retrieve Answer Key document URLs are valid and accessible.
5. Toggle the workflow to Active to enable production submissions.