pip install azure-ai-inference

from azure.ai.inference import ChatCompletionsClient
from azure.ai.inference.models import SystemMessage, UserMessage
from azure.core.credentials import AzureKeyCredential

client = ChatCompletionsClient(
    endpoint="<YOUR_TARGET_URI>",
    credential=AzureKeyCredential("<YOUR_KEY>"),
)

deploy_name = "<YOUR_DEPLOY_NAME>"

response = client.complete(
    messages=[
        SystemMessage(content="You are a helpful assistant."),
        UserMessage(
            content="Please think about ways to promote the use of generative AI in companies. Answer in Japanese." Switch the legend to Japanese,
        ),
    ],
    max_tokens=2048,
    model=deploy_name,
)

print(response.choices[0].message.content)

<think>
The first step is to clearly understand what the problem is. You need to think about how to promote the use of generative AI in the company.

Generative AI has a wide range of potential applications, so it's important to first assess and understand your company’s current situation. It's essential to identify which departments could benefit from AI adoption and what potential challenges they may face. For example, marketing, customer support, and business process automation are some potential areas.

Next, education and training are important. If employees are not provided with opportunities to learn how to use AI, it will be difficult to actually utilize it. Especially, to alleviate concerns and lack of understanding about AI, it might be a good idea to conduct workshops and training programs.

Selecting the right tools is also necessary. It is important to select a generative AI platform or tools that suit your company's needs Depending on the situation, customization may be necessary, and addressing security and data privacy issues is also required.

Support from the management is also essential. Without support from management, it will be challenging to secure the necessary budget and resources. Clearly communicating the benefits of AI implementation is essential to gaining management’s understanding and support.

Furthermore, making successful cases is effective. If certain departments successfully pilot AI and demonstrate tangible results, it becomes easier to scale adoption across other departments. Sharing the results company-wide and expanding on them can be an effective strategy.

It is also important to develop ethical guidelines and policies. Clear rules will need to be established to prevent the misuse of AI and encourage its responsible use.

Continuous evaluation and improvement also should not be overlooked. Even after implementing AI, it's crucial to evaluate its effectiveness and make improvements based on feedback to ensure ongoing progress.

It seems important to consolidate these points and consider concrete steps to move forward.
</think>

To promote the utilization of generative AI within a company, it is effective to systematically follow the steps below:

---

### **1. Current situation analysis and goal setting**
- **Identifying needs**: Identify the business challenges of each department (planning, sales, development, support, etc.) and clarify the areas where generative AI can provide solutions (document creation, data analysis, customer support, etc.).
- **Prioritization**: Prioritize the implementation of use cases with high impact and feasibility (e.g., automatic response to inquiries and automated report generation).

---

### **2. Building a system and changing awareness**
- **Launching a promotion team**: Establish a promotion team consisting of representatives from management, the IT department, and each department to collaborate to secure the budget and resources.
- **Implementation of educational programs**:  
  - **Basic understanding**: Share the mechanisms, benefits, and ethical risks (e.g., data leakage and bias) of generative AI through training.  
  - **Hands-on training**: Provide hands-on training to promote learning of prompt design and output verification methods.

---

### **3. Tool selection and test operation**
- **Choosing the right AI for your company**:  
  - Compare open sources (e.g., Llama 2) and cloud services (ChatGPT Enterprise, Google Vertex AI, etc.).  
  - Select based on security standards (e.g., data encryption and whether the pre-trained model can be customized).
- **PoC (Proof of Concept)**: Conduct test operations in a limited scope to verify accuracy and business efficiency improvement.

---

### **4. Establishment of governance**
- **Development of usage guidelines**:  
  - Set clear rules to prevent the input of confidential information and implement a system for double-checking output content.  
  - Address legal risks regarding copyright and privacy.
- **Monitoring System**: Regularly monitor the output quality and ethical aspects of AI and continuously improve them.

---

### **5. Cultivating a company culture**
- **Sharing successful cases**: Boost motivation by sharing success stories from departments that have achieved positive results, through internal presentations or reports.
- **Incentive design**: Recognize and reward teams that have improved business efficiency through the use of AI.

---

### **6. Collaboration with external entities and adaptation to evolution**
- **Collaboration with vendors and experts:**: Seek support for staying up to date with the latest technological trends and for implementing necessary customizations.
- **Continuous AI Updates**: Regularly review AI usage methods in response to advancements in model accuracy and changes in regulations.

---

### **Key points**
- **Management commitment**: Demonstrating return on investment (ROI) and securing management support with a long-term perspective is crucial.
- **Improvement driven by the field:**: Utilize feedback from employees (i.e., users) to optimize tools and processes.

Generative AI is a "tool," and the success or failure of its utilization depends on "how humans use it." In addition to introducing technology, improving digital literacy throughout the organization and ensuring psychological safety are keys.