Experience the Value of Agentiqs
Summary of This Page
You can experience the main value of Agentiqs in 90 minutes using the sample data.
- Experience knowledge utilization, meeting minutes generation, and design document review with a timetable
- Gain criteria for "what to check" at each step
- Includes purpose-based routes such as document search, review, and DENSO CREATE product integration
By reading this page, you can get a clear idea of whether Agentiqs can be used in your own business operations.
Overview
This page explains how to use the Agentiqs evaluation version to experience the main value of Agentiqs in 90 minutes.
Agentiqs is not just a tool for chatting with AI.
By combining AI agents, knowledge, workflows, file operations, and external tool integration, you can embed AI into your business processes.
In this 90-minute course, you will use sample data to quickly experience the following:
- Refer to business documents and receive answers
- Generate meeting minutes from meeting notes
- Review design documents based on review perspectives
- Output review findings in a structured format
- Standardize multiple processes by using workflows
- Understand how to apply these examples to your own business operations
If you have just obtained the evaluation version and want to know what to try first to understand the value of Agentiqs, follow the flow on this page.
What You Will Learn on This Page
This page explains the following:
- What you will experience in 90 minutes
- What to prepare before the evaluation
- A recommended 90-minute timetable
- Points to check in each sample
- What to look at to understand the difference from a typical AI chat
- How to think about replacing the samples with your own business tasks
- How to organize the evaluation results
- Which page to read next
Goals of This 90-Minute Course
The goal of this 90-minute course is to understand Agentiqs not as "just an AI chat," but as "a platform for embedding AI into business operations."
By the end of the 90 minutes, you should be able to explain the following.
| Goal | Description |
|---|---|
| Understand the basic value of Agentiqs | Understand the roles of AI agents, knowledge, workflows, and tool integration |
| Be able to run the samples | Run multiple representative samples and review their outputs |
| Understand the difference from a typical AI chat | Understand that it can work with business documents, review perspectives, tools, and output destinations |
| Find themes close to your own business tasks | Identify at least one business theme to try in your own organization |
| Decide the next evaluation step | Decide whether to move on to replacing with your own files, creating AI agents, or creating workflows |
The Value of Agentiqs You Will Experience in This 90-Minute Course
In this course, you will experience the following values in sequence.
| Value to Experience | Description | Sample Used |
|---|---|---|
| Knowledge utilization | AI refers to documents such as manuals and answers questions | Search across a large volume of documents |
| Deliverable generation | Generate meeting minutes from input such as meeting notes | Automatically generate meeting minutes |
| Review support | Review design documents based on review perspectives | Review a design document according to perspectives |
| Workflow automation | Standardize multiple processes and review by perspective | Rigorously review a design document by perspective |
| Application to your business | Understand how to replace samples with your own business tasks | Replace samples with your own business tasks |
Target Audience
This page is intended for the following users.
- Users who are using the Agentiqs evaluation version for the first time
- Users who want to quickly understand what Agentiqs can do
- Users who want to experience the value of AI agents and workflows
- Users who want to consider how AI can be applied to their own business operations
- Users considering use cases such as design, review, test design, and inquiry handling
- Users who want to confirm the value of integration with Next Design and Lightning Review
- Users who want to explain the evaluation results to their manager or team
What to Prepare in Advance
Before starting the 90-minute course, check the following.
| Check Item | Description |
|---|---|
| Starting Agentiqs | Confirm that you can start Agentiqs |
| AI provider settings | Confirm that you are able to chat with AI |
| Sample data placement | Confirm that the sample data is placed in the specified folder |
| Importing sample settings | Confirm that the sample AI agents, workflows, and tool settings have been imported |
| File access | Confirm that you can access the sample data input files and output destinations |
| Evaluation notes | Prepare notes to record what you notice during the evaluation |
If you have not yet checked the basic operations, review the following first.
Recommended Evaluation Environment
For your first evaluation, we recommend using the following environment.
| Item | Recommendation |
|---|---|
| Input data | Bundled sample data |
| Output destination | Verification folder under the sample data |
| Evaluation theme | Focus on one or two themes |
| Execution scope | Run the samples with their standard settings |
| External tools | Not required for the first evaluation. Try Next Design and Lightning Review later in the session or in a separate evaluation |
| Evaluator | A person who understands the business content and can judge the validity of the outputs |
For your first evaluation, do not use production files or production projects.
First confirm the behavior and value of Agentiqs with the bundled samples, and then move on to replacing them with your own files.
90-Minute Course Details
Overall Timetable
If you proceed in the following order, you can experience the main value of Agentiqs as a whole.
| Time | Activity | Purpose |
|---|---|---|
| 0–10 min | Prepare for the evaluation and confirm the goals | Decide what to focus on |
| 10–25 min | Experience knowledge utilization | Check how answers are generated by referring to business documents |
| 25–40 min | Experience meeting minutes generation | Generate deliverables from input data |
| 40–60 min | Experience design document review | Check how findings are extracted based on review perspectives |
| 60–75 min | Experience rigorous review with workflows | Standardize review by chapter and perspective |
| 75–90 min | Organize the evaluation results | Identify candidate use cases for your own business |
0–10 Minutes: Prepare for the Evaluation and Confirm the Goals
First, decide what you want to check during this 90-minute session.
Checkpoints
| Check Item | Description |
|---|---|
| Evaluation objective | What you want to confirm |
| Business areas of interest | Examples: document search, meeting minutes, design review, test design, and tool integration |
| Expected benefits | Examples: time savings, quality improvement, prevention of omissions, and deliverable generation |
| Evaluator | Whether the evaluator can judge the validity of the output |
| What to record | Note positive points, issues, and candidates for application in your own business |
Example Evaluation Notes
# Agentiqs 90-Minute Evaluation Notes
## Evaluation Objective
-
## Samples Tested
-
## Positive Points
-
## Points of Concern
-
## Themes That May Fit Our Business
-
## What to Try Next
-
10–25 Minutes: Experience Knowledge Utilization
First, try the sample for knowledge utilization.
Sample to Try
In this sample, a microcontroller manual is registered as knowledge, and the AI agent answers questions while referring to the manual.
What You Will Experience
In this sample, you will experience the following.
| Experience | What to Check |
|---|---|
| Knowledge reference | Whether the AI can answer by referring to the registered documents |
| Document search | Whether it can find the necessary information from a large volume of documents |
| Evidence-based answers | Whether the answers are based on the document content rather than general knowledge |
| Improving questions | Confirm that the answer quality changes depending on how the question is asked |
| Replacing with your own documents | Consider whether it could be used with your own manuals or specifications |
Example Questions to Try
Try questions like the following.
Please explain the initialization steps required to start using CAN communication.
Please organize the settings that should be checked when using receive interrupts in CAN communication.
Based on this manual, please create a checklist to use before implementing CAN communication.
Please summarize the constraints of CAN communication described in the manual in table format.
Do not guess any information that is not described in the manual. Reply with "Not stated" if it is not documented.
What are the recommended setting values for CAN communication?
Checkpoints
| Evaluation Perspective | What to Check |
|---|---|
| Based on the documents | Whether the answers follow the manual content |
| Not just general knowledge | Whether the answers are based on referenced knowledge |
| Useful in practice | Whether it seems likely to reduce investigation time |
| Easy to ask questions | Whether it is easy to ask follow-up or additional questions |
| Replaceable with your own documents | Whether it seems usable for product manuals, FAQs, or specifications |
Key Point to Understand at This Stage
The point you should confirm in this sample is as follows.
Agentiqs can answer by referring to business documents,
instead of answering based only on general AI knowledge.
If you want to use this in your own business, the following types of documents are good candidates.
- Product manuals
- Technical specifications
- Internal standards
- FAQs
- Past inquiries
- Design guides
- Review criteria
25–40 Minutes: Experience Meeting Minutes Generation
Next, try the sample for deliverable generation.
Sample to Try
In this sample, the AI agent generates meeting minutes from meeting notes as input.
What You Will Experience
In this sample, you will experience the following.
| Experience | What to Check |
|---|---|
| Reading input files | Whether the AI can read the meeting notes |
| Information organization | Whether it can organize decisions, ToDos, and discussion points |
| Applying an output format | Whether it can output in the specified meeting minutes format |
| Deliverable generation | Whether it can be used as a business deliverable rather than just a chat response |
| Replacing with your own notes | Whether it seems usable with your own meeting notes |
Example Requests to Try
Please create meeting minutes based on the attached meeting notes.
Please organize the following.
- Meeting overview
- Discussion details
- Decisions
- ToDos
- Open items
Please extract only the ToDos from the meeting notes.
Please organize them in the following table format.
| No | ToDo | Owner | Due Date | Notes |
Please organize the decisions made in this meeting separately from the items that should be checked next time.
Checkpoints
| Evaluation Perspective | What to Check |
|---|---|
| Information extraction | Whether it captures important information from the meeting notes |
| Structuring | Whether decisions, ToDos, and open items are separated |
| Appropriateness of supplementation | Whether it avoids making unfounded assertions about information not in the notes |
| Practical usability | Whether it can be used as a draft of meeting minutes |
| Ease of revision | Whether the format is easy for people to review and revise |
Key Point to Understand at This Stage
The point you should confirm in this sample is as follows.
Agentiqs can read business notes and input files
and generate deliverables used in actual work.
If you want to use this in your own business, the following use cases are good candidates.
- Creating meeting minutes
- Organizing ToDos
- Polishing meeting notes
- Creating drafts of reports
- Creating FAQ drafts
- Creating internal shared notes
40–60 Minutes: Experience Design Document Review
Next, try the sample for design document review.
Sample to Try
In this sample, the AI agent reviews a design document using the design document and review perspectives as input, and outputs identified issues.
What You Will Experience
In this sample, you will experience the following.
| Experience | What to Check |
|---|---|
| Reading the design document | Whether the AI can read the design document |
| Applying review perspectives | Whether it can check the document according to the specified perspectives |
| Extracting issues | Whether it can extract missing specifications, ambiguity, insufficient exception handling, and similar issues |
| Structuring findings | Whether it can organize the issue, reason, severity, and items to confirm |
| Replacing with your own design documents | Whether it seems usable for reviewing your own design documents |
Example Requests to Try
Please review the attached design document according to the attached review perspectives.
Be sure to organize the following.
- Issue
- Relevant section
- Reason for the issue
- Severity
- Items to confirm
- Improvement proposal
Please review the design document from the perspective of test design.
Point out missing conditions, expected results, boundary values, and exception cases needed to create test items.
Points to Check
| Evaluation Perspective | What to Check |
|---|---|
| Validity of the findings | Whether they are meaningful as actual review findings |
| Fit with the perspectives | Whether they follow the specified review perspectives |
| Evidence basis | Whether they are based on the content of the design document |
| Low level of guessing | Whether content not in the design document is being asserted |
| Reason for the finding | Whether it explains why this is a problem |
| Practical usability | Whether it seems usable for a self-check or first-pass review before the formal review |
Points to Understand at This Stage
The point you should confirm in this sample is as follows.
Agentiqs can check a design document using review perspectives
and create a draft of review findings.
If you want to use this in your own business, the following use cases are good candidates.
- Design document review
- Specification review
- API specification review
- Pre-check before test design
- Review support for junior staff
- Standardizing review perspectives
- Self-check before a formal review
60–75 Minutes: Experience Rigorous Review with a Workflow
Next, check the sample for rigorous review using a workflow.
Sample to Try
In this sample, the design document is split by chapter, and the AI review is run repeatedly for each combination of review perspectives.
What You Will Experience
In this sample, you will experience the following.
| Experience | What to Check |
|---|---|
| Workflow execution | Whether multiple steps can be run as a single series of processing |
| Chapter splitting | Whether the design document can be split and processed by chapter |
| Review by perspective | Whether the AI can review according to each review perspective |
| Coverage | Whether combining chapter x perspective seems to reduce missed checks |
| Output integration | Whether the results of multiple review runs can be consolidated into a list |
Image of the Processing
- Read the design document
- Split the design document by chapter
- Read the review perspectives
- Set the review target for each chapter
- Run the AI review for each review perspective, for each chapter
- Organize the issue, reason, severity, and items to confirm
- Consolidate all the review results
- Output them as a list of review findings
Points to Check
| Evaluation Perspective | What to Check |
|---|---|
| Standardization | Whether it can be processed without a person having to think through the steps each time |
| Coverage | Whether it can check using a combination of chapters and perspectives |
| Reproducibility | Whether it seems repeatable under the same conditions |
| Output quality | Whether the reviewer can easily check it as a list |
| Business applicability | Whether it seems usable for a first-pass check before the formal review |
Points to Understand at This Stage
The point you should confirm in this sample is as follows.
Agentiqs can standardize multiple processes as a workflow,
not just handle one-off requests to the AI.
If you want to use this in your own business, the following tasks are good candidates.
- Chapter-by-chapter review of design documents
- Multi-perspective review
- Test item generation
- Outputting a list of review findings
- Registering to an external tool
- Creating standard reports
75–90 Minutes: Organize the Evaluation Results
Finally, organize what you confirmed during the 90 minutes.
Perspectives for Organizing the Evaluation Results
| Item | What to Fill In |
|---|---|
| Samples tried | The names of the samples you ran |
| What was good | Points that felt usable in business |
| Points of concern | Concerns about output quality, operation, or settings |
| Theme close to your own business | The business task you want to try in your own organization |
| Input data to replace | Your own manuals, design documents, review perspectives, etc. |
| What to try next | Evaluating your own files, creating an AI agent, creating a workflow, etc. |
Evaluation Results Notes Template
# Agentiqs 90-Minute Evaluation Results
## Samples Run
- Search Across a Large Volume of Documents
- Automatically Generate Meeting Minutes
- Review a Design Document Based on Review Perspectives
- Rigorous Review of Design Documents by Review Points
## What Was Good
-
-
-
## Points of Concern
-
-
-
## Scenarios That May Be Usable in Your Own Business
| Theme | Why It Seems Usable | What to Check Next |
|---|---|---|
| | | |
## Candidates for Replacing with Your Own Files
| Candidate Data | Corresponding Sample | What You Want to Check |
|---|---|---|
| | | |
## Next Actions
- [ ] Try knowledge utilization with your own manuals
- [ ] Try review with your own design documents
- [ ] Replace the review perspectives with your own standards
- [ ] Create a review-specialized AI agent
- [ ] Adjust the workflow for your own business
Evaluation Points to Check at the End
At the end of the 90-minute course, check the following.
| Evaluation Point | What to Check |
|---|---|
| AI agent | Whether it seems usable as an AI tailored to a business purpose |
| Knowledge | Whether it seems able to refer to your own documents and answer |
| File input | Whether it seems able to read and process business files |
| Deliverable generation | Whether it seems able to create meeting minutes, lists of findings, test items, and similar deliverables |
| Review support | Whether it seems usable for pre-review checks or drafting candidate findings |
| Workflow | Whether it seems able to automate routine business tasks |
| Output format | Whether it seems able to output in a format that is easy for people to check and revise |
| Application to your organization | Whether you have an image of replacing it with your own data |
| Business impact | Whether it seems likely to lead to time savings, quality improvement, and prevention of omissions |
Points for Seeing the Difference from a Typical AI Chat
When evaluating Agentiqs, it is important to check the difference from a typical AI chat.
| Comparison Perspective | Typical AI Chat | Agentiqs |
|---|---|---|
| Using business documents | Tends to depend only on content pasted in on the spot | Can refer to knowledge and files |
| Role per business task | Needs to be specified in the prompt every time | Can be defined as the role of an AI agent |
| Routine processing | Requested manually every time | Can be turned into a workflow |
| Deliverable output | Mainly chat answers | Can output to Markdown, Excel, Word, and external tools |
| Review perspectives | Needs to be specified every time | Can be built into a perspective file or AI agent |
| External tool integration | Often involves manual transcription | Can integrate with tools such as Next Design and Lightning Review |
| Team rollout | Tends to become individual use | AI agents and workflows are easy to share |
What to Try if You Have Extra Time
If you have extra time within the 90 minutes, try one of the following.
Generate Test Design from a Design Document
What you can check:
- That test perspectives and test cases can be generated from a design document
- That you can experience cross-process support from design to test design
- That you get an image of replacing it with your own design documents
Extract Perspectives from Review Findings
What you can check:
- That past review findings can be reused
- That review perspectives or a checklist can be created
- Whether it seems usable for standardizing review quality
Try Next Design Integration
What you can check:
- That the AI can work with models in Next Design
- That test cases can be generated from a design model
- That model review can be supported by AI
Try Lightning Review Integration
What you can check:
- That AI-extracted review findings can be registered in Lightning Review
- Whether it seems able to reduce the work of transcribing into a review management tool
- That AI review results can be kept as a business deliverable
Arranging the 90 Minutes by Purpose
Instead of the standard route, you can also swap out the samples to match your purpose.
If You Prioritize Document Search and Inquiry Response
| Time | Content |
|---|---|
| 0–10 min | Evaluation preparation |
| 10–30 min | Search Across a Large Volume of Documents |
| 30–50 min | Consider replacing it with your own manuals or FAQs |
| 50–70 min | Check the page on utilizing knowledge |
| 70–90 min | Organize a policy for creating an AI for inquiry responses |
Recommended pages:
If You Prioritize Review Work
| Time | Content |
|---|---|
| 0–10 min | Evaluation preparation |
| 10–25 min | Extracting Perspectives from Review Findings |
| 25–45 min | Review a Design Document Based on Review Perspectives |
| 45–70 min | Rigorous Review of Design Documents by Review Points |
| 70–90 min | Organize a policy for creating a review-specialized AI agent |
Recommended pages:
- Extracting Perspectives from Review Findings
- Review a Design Document Based on Review Perspectives
- Rigorous Review of Design Documents by Review Points
- Creating an AI Agent Specialized for Reviews
If You Prioritize DENSO CREATE Product Integration
| Time | Content |
|---|---|
| 0–10 min | Evaluation preparation |
| 10–30 min | Generate Test Design from Design Documents |
| 30–55 min | Generate Test Cases for Each Requirement Model |
| 55–75 min | Review Next Design Models by Review Points |
| 75–90 min | Lightning Review integration, or organizing the evaluation results |
Recommended pages:
- Generate Test Design from Design Documents
- Generate Test Cases for Each Requirement Model
- Review Next Design Models by Review Points
- Automatically Register Review Findings to Lightning Review
How to Think About Replacing It with Your Own Business
Once you have confirmed the value of Agentiqs through the 90-minute course, the next step is to try replacing it with your own business.
At first, try not to substantially change the structure of the sample, and only replace the input data.
- Run the sample
- Check the sample's input data
- Choose one of your own data sets
- Replace only the input file
- Run it with the same AI agent or workflow
- Check the output results
- Adjust the perspectives or prompt as needed
Your Own Data That Is Easy to Replace
| Your Own Data | Corresponding Sample | What You Can Check |
|---|---|---|
| Product manual | Search Across a Large Volume of Documents | Document search, inquiry response |
| FAQ | Search Across a Large Volume of Documents | Drafting answers, FAQ search |
| Meeting notes | Automatically Generate Meeting Minutes | Creating meeting minutes, organizing ToDos |
| Design document | Review a Design Document Based on Review Perspectives | Extracting review findings |
| Design document | Generate Test Design from Design Documents | Generating test items |
| List of review findings | Extracting Perspectives from Review Findings | Creating review perspectives |
| Next Design project | Review Next Design Models by Review Points | Model review |
| Lightning Review file | Automatically Register Review Findings to Lightning Review | Registering findings |
Things to Avoid at First with Your Own Data
| Things to Avoid | Reason |
|---|---|
| Using a large volume of files at once | It becomes harder to isolate the cause if the output is poor |
| Mixing documents from multiple themes | It becomes harder for the AI to judge which information to refer to |
| Mixing old and new versions | The answer may be based on outdated information |
| Directly updating production files | Unintended output may remain |
| Registering directly to an external tool | You will need to check and delete the registered results |
| Evaluating with a theme whose correct answer is unknown | You cannot judge the validity of the AI's output |
Common Pitfalls and Solutions
| Pitfall | Common Cause | Solution |
|---|---|---|
| Not sure what to try first | There are many samples and it is hard to choose | Follow the standard route on this page |
| The answer becomes generic | The target knowledge or input file is not specified | Explicitly specify the target knowledge or input file |
| Too many findings | The review perspectives or target scope are too broad | Narrow the perspectives to 3-5 and the scope to one chapter or a few chapters |
| Too few findings | The review perspectives are vague, or the input information is insufficient | Make the perspectives more specific and check the content of the design document |
| It does not go well with your own files | There is a large structural difference from the sample | First replace only the input file, and leave other settings unchanged |
| The output format breaks down | The output conditions are vague | Explicitly specify the table format or column names |
| Stuck on external tool integration | Prerequisite settings or startup order are insufficient | Check the prerequisites on the sample page |
Notes for the Evaluation
- For your first try, run it with the bundled sample data.
- When trying it with your own files, use a copy made for verification.
- When handling confidential information, personal information, or customer information, follow your internal rules and security policies.
- Treat the AI's output as a draft for the person in charge to check, not as the final deliverable.
- When outputting or registering to an external tool, check the destination and the target to be updated in advance.
- Before outputting directly to a production file, production project, or production review, always confirm it with verification data first.
What to Check After Completing the 90-Minute Course
After completing the 90-minute course, check the following.
| Check Item | Description |
|---|---|
| Sample execution | Whether you were able to run the representative samples |
| Understanding the value | Whether you understood the difference between knowledge, deliverable generation, review, and workflows |
| Candidate for your own application | Whether you found a business theme to try in your own organization |
| Data to replace | Whether you decided on candidates such as your own manuals, design documents, or review perspectives |
| Next step | Whether you decided which step to take next: evaluating your own files, creating an AI agent, or creating a workflow |
| Internal explanation | Whether you can explain what was good, the points of concern, and what to try next |
Pages to Go to Next
Depending on your purpose, go to the following page.
| Purpose | Next Page |
|---|---|
| Check how to get started with the evaluation version | Evaluation Version Start Guide |
| Check basic operations | Quick Start |
| Check the sample data | Sample Data List |
| Try it with your own business files | Try with Your Own Business Files |
| Utilize knowledge | Utilizing Knowledge |
| Create an AI agent | Creating an AI Agent |
| Create a review-specialized AI agent | Creating an AI Agent Specialized for Reviews |
| Create a workflow | Creating a Workflow |
| Create a workflow for comprehensive review | Creating a Workflow for Comprehensive Review |
| Check the tool list | Tool List |