Experience DENSO CREATE Product Integration

Summary of This Page

You can integrate with Next Design, Lightning Review, and TimeTracker to experience cross-process support spanning design, review, and test design.

• Automatically generate test cases from requirement models (with Next Design traceability)
• Automatically register AI review findings to Lightning Review
• Includes recommended experience routes (60 minutes / 90 minutes / 120 minutes)

By reading this page, DENSO CREATE product users will understand where Agentiqs can be integrated into their work.

Overview

This page explains how to integrate Agentiqs with DENSO CREATE products (hereafter, DC products) and experience how AI can support tasks such as design, review, test design, and review finding management.

Agentiqs is not just a tool for chatting with AI.
By integrating with business tools such as Next Design, Lightning Review, and TimeTracker, the AI becomes able to work with design information, review findings, project information, and more.

In particular, integrating with Next Design or Lightning Review lets you experience the following kinds of use.

• The AI reads requirement models in Next Design
• It generates test cases for each requirement model
• The generated test cases are created as models in Next Design
• Traceability is set between the requirement models and the test case models
• The AI reviews design models in Next Design according to review perspectives
• Review findings extracted by the AI are registered in Lightning Review
• Design information, review findings, and downstream deliverables are connected and handled together

The purpose of this page is to let you quickly experience how, by integrating with DC products, Agentiqs can provide cross-process support spanning design, review, test design, and improvement activities.

What You Will Learn on This Page

This page explains the following:

• The value you can experience through DC product integration
• What you can do with Next Design integration
• What you can do with Lightning Review integration
• What you can do with TimeTracker integration
• The samples to try first
• Recommended routes for a 90- to 120-minute experience
• The points to check in each sample
• How to think about replacing samples with your own project
• What to check if things do not work well
• Which page to go to next

---

The Value You Can Experience Through DC Product Integration

With DC product integration, you can experience a flow in which Agentiqs treats information in your business tools as input and feeds the AI's output back into those tools.

Value You Can Experience	Description
Using design information with AI	The AI reads requirement models and design models in Next Design and uses them for review and deliverable generation
Generating deliverables across processes	Generating downstream deliverables such as test cases from requirement models or design documents
Traceability support	Working with relationships between deliverables, such as requirement models and test case models
Model review support	Checking design models in Next Design according to review perspectives
Managing review findings	Registering AI-extracted findings in Lightning Review to support review management
Reducing manual transcription work	Reflecting AI output in files and tools to reduce manual transcription
Embedding into business processes	Embedding AI into the flow of design, review, test design, and improvement activities

Overview of DC Product Integration

Combining Agentiqs with DC products lets you build a business flow such as the following.

1. Create design information in Next Design
2. Agentiqs retrieves the design models or requirement models
3. The AI agent understands the design information
4. Review findings or test cases are generated
5. The generated results are reflected in Next Design or Lightning Review
6. Review results and improvement information are accumulated
7. Project information from tools such as TimeTracker is combined in, as needed

This flow lets you embed the use of AI into the business process of design, review, and test design, rather than keeping it limited to one-off chats.

Products Covered by the Integration

This page mainly covers integration with the following products.

Product	Main Role	Example of Integration with Agentiqs
Next Design	System and software design support	Retrieving design models, model review, generating test cases, creating models
Lightning Review	Design review support	Retrieving review findings, registering AI findings, managing review results
TimeTracker	Work-hour and project management	Retrieving and using project information, work items, actual hours, and more

This page focuses mainly on samples for Next Design and Lightning Review.
TimeTracker integration requires a usage environment and API configuration, so check it if your environment is already set up.

What You Can Experience with Next Design Integration

With Next Design integration, you confirm that the AI can work with design models.

Main Things You Will Experience

Experience	Description
Model retrieval	Retrieving requirement models or design models in Next Design
Model understanding	The AI reads the model name, attributes, description, and related models
Test case generation	Generating test cases based on a requirement model
Model creation	Creating the generated test cases in Next Design
Setting traceability	Setting the relationship between the requirement model and the test case model
Model review	Checking the design model according to review perspectives
Outputting a list of findings	Listing the results of the model review

A Typical Processing Flow for Next Design Integration

1. Open the Next Design project
2. Select the model to be reviewed or used as the basis for generation
3. Agentiqs retrieves the model information
4. The AI agent understands the model content
5. Review findings or test cases are generated
6. A model is created in Next Design, if needed
7. Check the results

Prerequisites for Trying Next Design Integration

Before trying Next Design integration, check the following.

Check Item	Description
Starting Next Design	You can start Next Design
Starting Agentiqs	Agentiqs is running
Sample project	You can open the sample Next Design project
Integration settings	Agentiqs and Next Design are in a state where they can be integrated
Target model	You can check the requirement models or design models
Data for verification	Use the sample or a copied project for your first try

Things to Avoid at First with Next Design Integration

Things to Avoid	Reason
Directly using the production project	Unnecessary models or changes may remain
Selecting a large number of target models	It becomes harder to check the results and isolate causes
Changing the metamodel from the start	The integration processing becomes complex and harder to verify
Making major changes to the traceability settings	It becomes harder to check the generated results
Targeting multiple types of models at once	The AI's output and processing conditions become complex

---

What You Can Experience with Lightning Review Integration

With Lightning Review integration, you experience the flow of registering AI-extracted review findings into a review management tool.

Main Things You Will Experience

Experience	Description
Finding registration	Registering AI-extracted findings in Lightning Review
Structuring findings	Organizing the issue, relevant section, reason, severity, and items to confirm
Review management	Managing the output as review findings rather than as chat answers
Reducing transcription	Reducing the burden of manually transcribing AI output
Accumulating review results	Keeping the findings in the review management tool

A Typical Processing Flow for Lightning Review Integration

1. The AI checks the design document or the item to be reviewed
2. It extracts findings according to the review perspectives
3. The issue, reason, severity, and items to confirm are organized
4. They are converted into a format for registration in Lightning Review
5. The findings are registered in Lightning Review
6. The reviewer checks them in Lightning Review

Prerequisites for Trying Lightning Review Integration

Before trying Lightning Review integration, check the following.

Check Item	Description
Starting Lightning Review	You can start Lightning Review
Starting Agentiqs	Agentiqs is running
Sample review	You can open the sample or a review for verification
Write permission	You are able to register findings
Checking the destination	The review you are registering findings to is correct
Data for verification	Use data for verification rather than the production review for your first try

Things to Avoid at First with Lightning Review Integration

Things to Avoid	Reason
Registering directly to the production review	Unnecessary findings may remain
Registering a large number of findings in bulk	It becomes difficult to check and delete them
Running it without checking the destination review	Findings may be registered to an unintended review
Registering without checking the content of the findings	Inappropriate findings may end up being managed
Not checking for duplicates	The same finding may be registered multiple times

---

What You Can Experience with TimeTracker Integration

With TimeTracker integration, Agentiqs becomes able to work with projects, work items, actual hours, workload data, and more.

If the TimeTracker API settings have been completed, the following kinds of use become possible.

What You Can Experience	Description
Retrieving project information	Retrieving the project list and detailed information
Retrieving work items	Retrieving work items, tasks, and planning information
Checking actual hours	Retrieving actual hours and using them to understand the status
Checking workload data	Retrieving workload by member or by period
Organizing the status	The AI summarizes the retrieved information and organizes the status and items to confirm
Considering improvements	Organizing risks and improvement points based on hours and progress information

An Image of Using TimeTracker Integration

1. Retrieve project information from TimeTracker
2. Retrieve work items and actual hours
3. The AI summarizes the status
4. Delay risks, workload imbalances, and items to check are organized
5. A report or improvement memo is created, if needed

Prerequisites for Trying TimeTracker Integration

Check Item	Description
TimeTracker environment	You can connect to the TimeTracker server
API settings	The server URL and API key are configured
Permissions	You have permission to retrieve the target project and actual hours
Target project	You can specify the project to use for evaluation
Verification scope	Limit it to a small number of projects or a short period for your first try

---

Samples to Try First

If this is your first time experiencing DC product integration, we recommend proceeding in the following order.

Order	Sample	What You Can Experience
1	Generate Test Design from Design Documents	Experience the document-based flow from design to test design
2	Generate Test Cases for Each Requirement Model	Generate test cases from a Next Design requirement model
3	Review Next Design Models by Review Points	Have the AI review a Next Design model
4	Automatically Register Review Findings to Lightning Review	Register AI-extracted review findings in Lightning Review

If you do not yet have a Next Design or Lightning Review environment, start by trying "Generate Test Design from Design Documents" to experience cross-process deliverable generation.
After that, moving on to the Next Design and Lightning Review integration samples will be easier to understand.

Recommended Experience Routes

Choose one of the following experience routes depending on your purpose.

Purpose	Recommended Route
You want to quickly get a feel for it in a short time	60-minute route
You want to thoroughly check Next Design integration	90-minute route
You want to check Next Design and Lightning Review all the way through	120-minute route
You want to check whether it can be used with your own project	Your own project replacement route

---

60-Minute Route: Experience the Value of Cross-Process Support in a Short Time

This route is for those who want to experience Agentiqs' cross-process support first, without spending time preparing a Next Design or Lightning Review environment.

1. Generate test design from a design document
2. Experience design document review
3. Check what is possible with DC product integration

Samples to Try

Order	Sample	Purpose
1	Generate Test Design from Design Documents	Check the flow of generating test design from design information
2	Review a Design Document Based on Review Perspectives	Check the design document according to review perspectives

What to Check in This Route

Check Perspective	Description
Using design information	Whether the AI can read the design document and convert it into downstream deliverables
Test design support	Whether it can create a draft of test perspectives or test cases
Review support	Whether it can point out missing specifications or ambiguity in the design document
Image of applying it to your own organization	Whether the same flow seems triable with your own design documents

---

90-Minute Route: Experience Next Design Integration

This route is for those who mainly want to check Next Design integration.

1. Generate test design from a design document
2. Generate test cases for each requirement model
3. Review Next Design models by perspective
4. Organize the evaluation results

Samples to Try

Order	Sample	Purpose
1	Generate Test Design from Design Documents	Understand document-based test design generation
2	Generate Test Cases for Each Requirement Model	Generate test cases from a Next Design requirement model
3	Review Next Design Models by Review Points	Experience reviewing a Next Design model

What to Check in This Route

Check Perspective	Description
Model retrieval	Whether Agentiqs can retrieve the model in Next Design
Model understanding	Whether the AI can work with the model name, attributes, description, and relations
Deliverable generation	Whether test cases can be generated from the requirement model
Model creation	Whether the generated information can be created as a model in Next Design
Traceability	Whether the relationship between the requirement model and the test case model can be handled
Model review	Whether the design model can be checked according to review perspectives

---

120-Minute Route: Experience Next Design and Lightning Review Integration

This route is for those who want to check both Next Design and Lightning Review, and experience the flow from design through to review finding management.

1. Generate test cases for each requirement model
2. Review Next Design models by perspective
3. Check the findings from the design document review or model review
4. Register the review findings in Lightning Review
5. Organize the evaluation results

Samples to Try

Order	Sample	Purpose
1	Generate Test Cases for Each Requirement Model	Generate test cases from a Next Design requirement model
2	Review Next Design Models by Review Points	Have the AI review a Next Design model
3	Automatically Register Review Findings to Lightning Review	Register AI-extracted review findings in Lightning Review

What to Check in This Route

Check Perspective	Description
Using design models	Whether the AI can work with design information in Next Design
Review support	Whether review perspectives can be applied to the model
Structuring findings	Whether the issue, target model, reason, and severity can be organized
Registering to Lightning Review	Whether AI-extracted findings can be registered in the review management tool
Turning it into a business process	Whether AI seems embeddable into the flow of design, review, and finding management

---

Details by Experience Route

From here, this section explains the points to check in each sample.

1. Generate Test Design from Design Documents

Sample

• Generate Test Design from Design Documents

What You Will Experience in This Sample

Using the design document as input, the AI generates a draft of test perspectives and test cases.

Before moving on to Next Design integration, you can first check the document-based flow of "generating test design from design information."

Processing Flow

1. Read the design document
2. Read the test perspectives
3. The AI understands the design content
4. Generate the test design according to the test perspectives
5. Check the generated results

Points to Check

Check Perspective	Description
Design understanding	Whether the inputs, outputs, conditions, and constraints of the design document are read correctly
Applying test perspectives	Whether the output is generated according to the specified test perspectives
Expected results	Whether the expected results are based on the design document
Unclear points	Whether gaps in the design document are output as items to confirm
Application to your organization	Whether it seems triable with your own design documents

---

2. Generate Test Cases for Each Requirement Model

Sample

• Generate Test Cases for Each Requirement Model

What You Will Experience in This Sample

Using a requirement model in Next Design as input, the AI generates test cases for each requirement.
The generated test cases are created as test design models in Next Design, and traceability is set with the requirement models.

Processing Flow

1. Open the Next Design project
2. Retrieve the requirement model
3. The AI understands the requirement content
4. Generate test cases for each requirement
5. Create the test case models in Next Design
6. Associate the requirement models with the test case models

Points to Check

Check Perspective	Description
Model retrieval	Whether the requirement model in Next Design can be retrieved
Requirement understanding	Whether the AI reads the content of the requirement model correctly
Test case generation	Whether valid test cases are generated for each requirement
Model creation	Whether the test case model is created in Next Design
Traceability	Whether the relationship between the requirement model and the test case model is set
Application to your organization	Whether it seems replaceable with your own requirement models or metamodel

When Trying It with Your Own Project

For your first try, use the following conditions.

Item	Recommendation
Project	A copied project for verification, not the production project
Target model	1-5 requirement models
Test perspectives	3-5 perspectives
Output destination	A test case model for verification
Evaluation method	Have a person check the generated model

---

3. Review Next Design Models by Review Points

Sample

• Review Next Design Models by Review Points

What You Will Experience in This Sample

Using a design model in Next Design as input, the AI agent runs a review for each review perspective.

By checking the design model on a "model x review perspective" basis, you can experience model-based AI review, which differs from document review.

Processing Flow

1. Open the Next Design project
2. Retrieve the model to be reviewed
3. Read the review perspectives
4. Organize the information to be reviewed for each target model
5. Run the AI review for each model, by review perspective
6. Organize the issue, reason, severity, and items to confirm
7. Output a list of review findings

Points to Check

Check Perspective	Description
Model understanding	Whether the AI can work with the model name, attributes, description, and relations
Applying the perspectives	Whether the check follows the specified review perspectives
Validity of the findings	Whether the findings are meaningful as a model review
Evidence basis	Whether the findings are based on the model information
Output format	Whether the issue, reason, and severity are easy to check
Application to your organization	Whether it seems adaptable to your own metamodel or review criteria

When Trying It with Your Own Project

Item	Recommendation
Project	A copied project for verification
Target model type	Limit it to one type
Number of target models	1-5
Review perspectives	3-5 perspectives
Output destination	An Excel or Markdown file for verification
Evaluation method	Have the design owner or reviewer check it

---

4. Automatically Register Review Findings to Lightning Review

Sample

• Automatically Register Review Findings to Lightning Review

What You Will Experience in This Sample

AI-extracted review findings are automatically registered in Lightning Review.

You can check a flow where the AI's output does not stop at chat or a file, but is left as a finding in the review management tool.

Processing Flow

1. Read the design document
2. Read the review perspectives
3. The AI extracts the review findings
4. Organize the issue, reason, severity, and items to confirm
5. Convert them into the Lightning Review registration format
6. Register the findings in Lightning Review
7. Check the registration results

Points to Check

Check Perspective	Description
Finding registration	Whether the findings are registered in Lightning Review
Quality of the findings	Whether the findings are meaningful in practice
Registered items	Whether the issue, relevant section, reason, and severity are easy to see
Duplicates	Whether there are unnecessary duplicate findings
Reducing transcription	Whether it seems likely to reduce manual registration of findings
Application to your organization	Whether it seems adaptable to your own review operations

When Trying It with Your Own Review

Item	Recommendation
Lightning Review file	A copy or a review for verification, not the production review
Target design document	1 file
Scope	1-3 chapters
Review perspectives	3-5 perspectives
Number of registrations	Up to about 10
Evaluation method	Have the reviewer check the registration results

When Replacing with Your Own Project

When replacing the DC product integration samples with your own project, proceed carefully.

In particular, Next Design and Lightning Review may create models or register findings against your actual project or review.
For your first try, always use copied data for verification.

Basic Steps for Replacing with Your Own Project

1. Confirm normal operation with the sample data
2. Copy your own project or review
3. Narrow the scope down to something small
4. Replace only the input data
5. Specify an output destination for verification
6. Run it
7. Check the output results or registration results
8. Adjust the perspectives or prompts, as needed

What to Change First When Replacing with Your Own Project

What to Change	Example
Target model	Requirement models, function models, state models, etc.
Target design document	A design document copied for verification
Review perspectives	Your own review criteria, quality standards
Test perspectives	Your own test design standards
Output destination	A verification Excel file, verification model, or verification review
Scope	1-5 models, 1-3 chapters, 3-5 perspectives

What Not to Change at First When Replacing with Your Own Project

What Not to Change	Reason
Production project	Unnecessary models or findings may remain
Metamodel structure	Changing it from the start makes it harder to isolate causes
Overall workflow structure	It becomes harder to identify the cause if it stops working
External tool registration processing	Misconfiguration of the destination or update target is more likely
Large-volume data processing	Checking and deleting the results becomes difficult
Basic AI agent prompt	The output quality may change significantly

---

Evaluation Perspectives for DC Product Integration

When evaluating DC product integration, check it from the following perspectives.

Evaluation Perspective	What to Check
Tool integration	Whether it can integrate with Next Design or Lightning Review
Input retrieval	Whether design models, design documents, and review information are retrieved correctly
AI understanding	Whether the AI understands the model or design information
Deliverable generation	Whether test cases, review findings, and similar deliverables can be generated
Reflecting the output	Whether the generated results can be reflected in Next Design or Lightning Review
Traceability	Whether relationships such as those between requirements and test cases can be handled
Practical usability	Whether it seems usable for design, review, and test design tasks
Safety	Whether an operation that avoids accidentally updating production data seems achievable
Business impact	Whether it seems likely to lead to reduced transcription, more efficient reviews, and improved quality

---

Points for Seeing the Difference from a Typical AI Chat

With DC product integration, it becomes especially easy to confirm the difference from a typical AI chat.

Comparison Perspective	Typical AI Chat	Agentiqs + DC Product Integration
Handling design information	A person needs to paste in the design content	Can retrieve the model in Next Design
Deliverable generation	Answers as text	Can be created as a model in Next Design
Traceability	Managed manually	Can handle the relationship between requirements and test cases
Review findings	Output in the chat	Can be registered in Lightning Review
Routine processing	Requested manually every time	Can be run as a workflow
Business process	Tends to stay limited to individual work	Can be embedded into the flow of design, review, and test design

---

Evaluation Results Notes Template

After experiencing DC product integration, organize the evaluation results as follows.

DC Product Integration Evaluation Notes

# DC Product Integration Evaluation Notes

## Samples Tried
- 

## Products Used
- [ ] Next Design
- [ ] Lightning Review
- [ ] TimeTracker

## Data Used
- Next Design project:
- Lightning Review file:
- Design document:
- Review perspectives:

## What Was Confirmed
-
-
-

## What Was Good
-
-
-

## Points of Concern
-
-
-

## Scenarios That May Be Usable in Your Own Business

| Business Scenario | How to Use It | Expected Benefit |
|---|---|---|
|  |  |  |

## What to Check Next
- [ ] Try it with your own Next Design project
- [ ] Replace it with your own review perspectives
- [ ] Register findings to a Lightning Review for verification
- [ ] Adjust the AI agent for your own organization
- [ ] Adjust the workflow for your own business
- [ ] Evaluate it as a team

## Challenges Toward Adoption
-
-
-

---

Common Pitfalls and Solutions

Pitfall	Common Cause	Solution
Cannot integrate with Next Design	The startup order or extension settings are insufficient	Start or restart Next Design after starting Agentiqs
Cannot retrieve the model	The target model is not selected, or the settings do not match	Check the target model, metamodel, and view
The generated model differs from expectations	The description of the requirement model or the test perspectives are insufficient	Make the description of the target model and the test perspectives more specific
Traceability is not set	The association target or metamodel settings do not match	Check the difference between the sample's structure and your own metamodel
The model review findings are generic	The review perspectives are abstract, or the model information is insufficient	Make the perspectives more specific and enrich the model attributes and description
Not registered in Lightning Review	There is a problem with the destination review, the startup state, or permissions	Check the destination and the write permission
Too many findings	The scope or perspectives are too broad	Narrow down the number of target models or perspectives
Worried about updating production data	A copy for verification has not been prepared	Always run it with a copied project or review

---

How to Proceed When Applying It to Your Own Organization

When applying DC product integration to your own business, we recommend proceeding in the following order.

1. Confirm normal operation with the sample
2. Prepare your own data for verification
3. Narrow the scope down to something small
4. Replace only the input data
5. Have a person check the output results
6. Align the review perspectives or test perspectives with your own standards
7. Adjust the AI agent
8. Adjust the workflow
9. Try it on a small scale within your team

Tasks Suited to Applying It in Your Own Organization

Task	How to Use Agentiqs + DC Product Integration
Requirement review	Check the requirement model in Next Design according to review perspectives
Design review	Have the AI review the design model or design document
Test design	Generate test cases from the requirement model or design document
Review finding management	Register AI findings in Lightning Review
Creating review perspectives	Extract review perspectives from past findings
Checking traceability	Check the relationships between requirements, design, and tests
Improvement activities	Analyze finding trends and weaknesses in design quality

The Business Image Targeted by DC Product Integration

The goal of DC product integration is not to use AI as a one-off aid.
It is to connect design information, review results, test design, and project information, and support the business process as a whole.

1. Design
2. The AI understands the design information
3. Review
4. Manage the findings
5. Design the tests
6. Manage the relationships between deliverables
7. Use it for improvement activities

Within this flow, Agentiqs supports the work of designers, reviewers, test designers, and project managers.

---

Notes

• Always confirm the behavior with the sample data first.
• When trying it with your own project, always use data copied for verification.
• When outputting or registering to Next Design or Lightning Review, check the target project, the destination review, and the output destination in advance.
• Before outputting directly to a production project or production review, always confirm it with verification data first.
• Treat the AI's output as a draft for the person in charge to review, not as the final deliverable.
• Have the person in charge check the generated models, review findings, and test cases before using them.
• When handling confidential information, personal information, or customer information, follow your internal rules and security policies.
• For tools that require API integration, such as TimeTracker, GitHub, Jira, and Confluence, manage your API keys and tokens appropriately.

---

What to Check After the Experience

After experiencing DC product integration, check the following.

Check Item	Description
Next Design integration	Whether you were able to confirm model retrieval, model creation, and traceability settings
Lightning Review integration	Whether you were able to register AI findings in the review management tool
TimeTracker integration	Whether it seems possible to work with project information and hours information, if needed
Output quality	Whether the generated test cases and review findings seem usable in practice
Applicability to your organization	Whether it seems adaptable to your own project, review perspectives, and metamodel
Business impact	Whether it seems likely to lead to reduced transcription, more efficient reviews, and improved quality
Next verification	Whether you decided which business theme to continue evaluating

---

Pages to Go to Next

Depending on your purpose, go to the following page.

Purpose	Next Page
Return to the sample data list	Sample Data List
Check how to get started with the evaluation version	Evaluation Version Start Guide
Experience the value of Agentiqs in 90 minutes	Experience the Value of Agentiqs in 90 Minutes
Try it with your own business files	Try with Your Own Business Files: 90-Minute Course
Generate test design from a design document	Generate Test Design from Design Documents
Generate test cases from a Next Design requirement model	Generate Test Cases for Each Requirement Model
Review a Next Design model	Review Next Design Models by Review Points
Register findings to Lightning Review	Automatically Register Review Findings to Lightning Review
Check the tool list	Tool List
Create a workflow	Creating a Workflow