Model HQ
DocumentationBuild a No-Code Document Review and Analysis Custom Agent Workflow in Model HQ
Create powerful document processing agents without writing any code
The No-Code Document Review and Analysis Workflow in Model HQ empowers business users and analysts to build custom document processing agents—without writing a single line of code. Using a visual editor and customizable templates like the Contract Analyzer, users can tailor workflows to fit specific use cases, such as reviewing music license royalty agreements. From editing questions to adding natural language summaries, this workflow enables fast, scalable, and domain-specific analysis that can be tested on single documents or run in batch across entire folders—ready for sharing or backend integration.
Video Tutorial Available
This walkthrough is also demonstrated step-by-step on our YouTube video:
Build a Custom Chat/RAG Bot in 10 MinutesUse Case
Custom analysis of music license royalty agreements as an example, derived from an existing Contract Analyzer agent.
Who This Is For
Business Users
HR, Legal, Operations teams
Developers
Building custom workflows
Analysts
Automating document review
Ingredients (Prerequisites)
- Model HQ installed
- A sample agent (e.g., Contract Analyzer) available as a template in Model HQ
- A sample PDF — e.g., Circe License Agreement for Copyrighted Music - sample document provided as part of Model HQ.
C:\users\[username]\llmware_data\sample_files\AgreementsLarge\Circe License Agreement for Copyrighted Music.pdf
- No coding knowledge required!
How-To Covered
Use the no-code visual editor to:
- Add/remove rows
- Update question text
- Reuse the same RAG answer service block
Step-by-Step Recipe
Start a New Agent Workflow
Goal: Create a new custom agent by cloning an existing one.
Steps:
- 1Open Agents → Select Build New then click >
- 2Next, select Start Building in Configure Agent
- 3Input process name in Set Up Agent, e.g., License Agreement
- 4Set input type to User Document (indicates PDF upload)
- 5Derive from existing agent: select Contract Analyzer

Note
The input type selection is critical in Model HQ—determining how the agent receives input. Multiple input sources can be configured in advanced use cases.
Understand the Template Process (Contract Analyzer)
You'll see pre-loaded steps for this template which was designed for employment agreements, which we will modify for the new contracts related to music license royalty agreements:
- 1Parse the uploaded document
- 2Run RAG answer agent on 3 default questions:
- When is the effective date?
- What is the annual rate of the base salary?
- How many vacation days?
- 3Output results to a report
Customize the Workflow for New Use Case
Scenario: We're now analyzing music license royalty agreements, not employment contracts.
Keep:
- Parse Document (critical for extracting text from non-
.txt
files)
Modify Questions:
- Change 2nd row to: "When is the effective date of the agreement?"
- Change 3rd row to: "Who are the parties to the agreement?"
- Change 4th row to: "What are the royalty terms, including amount and calculation method?"
Enhance Output: Add a Commentary Step
Optional Twist: Replace the basic report with a smarter summary.
Steps:
- Add a new block: Report Commentary
- Pass in all previous answers
- Skip generating Word reports—commentary replaces it
- Output: Natural language summary of findings
How to in the workflow:
- 1At the end of Row 5, click the '+' button
- 2Select Report Commentary in the Service dropdown
- 3Input:
Natural Language Summary of Findings
- 4Context: None

- Click
>
thenRun
then in Action click>
- Upload the document:
C:\users\[username]\llmware_data\sample_files\AgreementsLarge\Circe License Agreement for Copyrighted Music.pdf

Test the Agent on a Single Document
Steps:
- Click Run
- Upload a music license agreement PDF
Model HQ will:
- Parse the document (e.g., ~17 text chunks)
- Run the RAG agent for all 4 custom questions
- Generate a clean summary commentary of the findings

Batch Run the Agent on Multiple Documents
Steps:
- Switch to Batch Mode:Go to Agents → Load Existing → License Agreement (your custom agent), then select Batch Run
- Select 3–4 music license agreements, e.g.:
- Circe License Agreement for Copyrighted Music
- Cybele Music License Agreement
- Eos License Agreement for Copyrighted Music
Folder path: C:\users\[username]\llmware_data\sample_files\AgreementsLarge\

In Upload Documents for Batch Processing, after selecting documents, click >
to begin the batch process
Model HQ will:
- Run the agent on each file
- Output per document:
- Question-answer results
- One commentary summary
- Ready for export or further analysis

(Optional) Share or Integrate
You can now:
- Email the agent to a colleague (they can upload & run it)
- Save it to a shared Model HQ repository
- Deploy it via Model HQ API Server to integrate into an app
- Expose the workflow programmatically to backend services
Summary Table
Step | Action |
---|---|
Derive New Agent | Based on Contract Analyzer |
Modify Questions | Remove irrelevant, add domain-specific |
Add Commentary | Auto-generate summary from answers |
Run Single Test | Validate logic and structure |
Run Batch Test | Scale to many docs at once |
Share / Integrate | Via file or backend API |
Why It Works
Fast Adaptation
Using no-code UI for rapid development
Domain-Agnostic
Works for any document type
Quick Deployment
Test, tweak, and deploy in minutes
User Empowerment
Empowers analysts and business users without developer time
Need Help?
If you encounter any issues while updating your model configuration, feel free to contact our support team at support@aibloks.com