Welcome to the AI Assist for Jira Service Management FAQ page! Here, you'll find a collection of frequently asked questions (FAQs) that are categorized for easy reference.

What is AI?

AI stands for Artificial Intelligence, which is a branch of computer science that focuses on creating intelligent machines that can perform tasks that normally require human intelligence, such as learning, problem-solving, decision-making, and understanding natural language.

What is NLP?

NLP stands for Natural Language Processing, which is a subfield of AI that deals with the interaction between computers and human languages. It involves teaching machines to understand, interpret, and generate human language, such as text or speech. NLP is a key technology behind chatbots, language translation, and voice assistants like Siri or Alexa.

Which AI service does AI Assist for Jira Service Management use?

The app AI Assist for Confluence uses an AI service called OpenAI, which is sometimes referred to as ChatGPT. This service is a type of artificial intelligence that specializes in natural language processing, meaning it can understand and generate human language.

What is OpenAI?

OpenAI is a research organization that focuses on developing advanced AI technologies for the benefit of humanity. Their work spans a range of AI applications, including NLP, robotics, computer vision, and game-playing. They are known for their breakthroughs in natural language processing, such as the GPT series of language models that can generate coherent and human-like text.

What is AI Assist for Jira Service Management?

AI Assist for Jira Service Management is an app that integrates OpenAI's ChatGPT with your Jira Service Management projects, helping agents resolve customer requests more efficiently.

How do I enable AI Assist for my Jira Service Management project?

To enable AI Assist, enter a valid OpenAI API key in the project's settings. Note that a paid OpenAI subscription is required. Once configured, agents can start using AI Assist and enable the Auto Suggest feature.

How do I request AI assistance for a customer request?

Within a customer request, click on the "AI Assist: Request AI Assistance" option in the left sidebar. Customize your AI Assist request by selecting the fields to include and providing additional context or instructions as needed.

What is the Auto Suggest feature?

The Auto Suggest feature, when enabled, automatically adds an internal comment with an AI-generated solution and/or workaround to each new customer request, helping agents resolve requests more quickly.

What are the available settings in the AI Assist Settings page?

The AI Assist Settings page includes options to configure the OpenAI API key and toggle the Auto Suggest feature on or off. A metrics tab provides insights into AI assist usage, request costs, and a detailed overview of the last 1,000 AI assists.

Why does ChatGPT sometimes generate an inaccurate or completely made-up solution to a request?

ChatGPT, although powerful, is an AI model and not a human expert. It can occasionally produce inaccurate or made-up solutions based on the data it was trained on or the way it interprets the input. The model's understanding of context and real-world knowledge might be imperfect, leading to suboptimal or incorrect responses. It's crucial for users to review AI-generated solutions before taking action or sharing them with customers.

Why doesn't AI Assist provide an option for Auto Suggest to share the AI suggestion directly with the customer?

AppsFoundry prioritizes delivering accurate and professional support to customers and their customers. Since AI-generated suggestions can occasionally be inaccurate or incomplete, Auto Suggest is designed to add suggestions as internal comments first. This allows agents to review, edit, and verify the AI-generated responses before sharing them with customers, ensuring that the information provided is of high quality and meets customer expectations.

Are there any limitations to the AI Assist requests?

Yes, the AI Assist requests are subject to limitations imposed by OpenAI's API, including rate limits and token limits. Please refer to OpenAI's documentation for more information on these limitations.

Does AI Assist for Jira Service Management support all languages?

While ChatGPT is primarily trained on English content, it may provide some support for other languages. However, the quality and accuracy of responses in non-English languages may vary and are not guaranteed.

Can AI Assist handle highly technical or specialized requests?

Although AI Assist can often provide useful suggestions for a wide range of topics, its understanding of highly specialized or technical domains may be limited. It's essential for agents to review and verify AI-generated responses before sharing them with customers.

Is there a way to improve AI Assist's understanding of my specific domain or industry?

Currently, AI Assist for Jira Service Management doesn't support custom training or fine-tuning for specific domains or industries. However, providing additional context or instructions in the AI Assist request may help improve the quality and relevance of the generated responses.

Can I use AI Assist for Jira Service Management without a paid OpenAI subscription?

No, a paid (pay-as-you-go) OpenAI subscription is required to use AI Assist for Jira Service Management. You will need to enter a valid OpenAI API key in your project settings to enable the AI Assist features.

Does AI Assist for Jira Service Management store any sensitive (GDPR) data?

No, AI Assist for Jira Service Management does not store any sensitive or GDPR-related data. The only data stored are the encrypted OpenAI API key and usage metrics, which do not contain any personal data.

Where is the data stored?

The data is stored within Atlassian's cloud infrastructure and not on AppsFoundry's systems. Data processing also takes place within Atlassian's cloud, except for the data sent to OpenAI, which is processed by OpenAI.

How is the OpenAI API key stored?

The OpenAI API key is stored encrypted within Atlassian's cloud infrastructure, ensuring its security and privacy.

Is AppsFoundry responsible for the type of data sent to OpenAI?

As per the app's EULA, AppsFoundry is not responsible for the type of data sent to OpenAI. Users are responsible for ensuring that the data sent complies with applicable regulations and privacy requirements.

Where can I find more information about OpenAI's terms of use and Data Security & Privacy Statement?

You can find OpenAI's terms of use at https://platform.openai.com/docs/terms-of-service and their Data Usage Policy at https://platform.openai.com/docs/data-usage-policy .

How can I ensure the data I send to OpenAI is secure and compliant with privacy regulations?

Users are responsible for adhering to data security and privacy regulations when sending data to OpenAI. Review the data sent to OpenAI to ensure it does not contain sensitive, intellectual property, or GDPR-related information. Always follow your organization's data security and privacy guidelines.

Why does AppsFoundry require individual OpenAI API keys for AI Assist for Jira Service Management?

AppsFoundry requires individual OpenAI API keys to prioritize data privacy, security, and customer control. Using individual API keys ensures that you have full control over your data and costs while minimizing the risk of data breaches or unauthorized access. To learn more about our approach and its benefits, please refer to our article on why we chose individual OpenAI API keys.

Is payment required to use OpenAI?

Yes, OpenAI offers a range of APIs and services for developers to utilize, and some of these may require payment to access. However, OpenAI also provides a free research preview service, which allows developers to experiment with cutting-edge AI models and technologies in a limited capacity. This service is designed for research and experimentation purposes and may not have the same level of support and reliability as the paid APIs and services. Additionally, some of the paid APIs may have free usage tiers or limited free trials available, allowing developers to test out the features before committing to a paid plan.

Overall, OpenAI offers a range of pricing plans based on usage and specific features, and users can choose the plan that best fits their needs.

Useful links

• OpenAI - Pricing (we use the GPT-3.5 Turbo model)

• Signup for an OpenAI account

What is the cost of a single AI assist?

In short, the cost depends on the data sent to ChatGPT and the response received, making it challenging to predict the exact cost of each request. Generally, more detailed requests are costlier. However, from our observations, the average AI assist falls between $0.001 and $0.002. This translates to up to 1000 AI assists for just $1 – an affordable option compared to competitors that bundle OpenAI subscriptions, where you may pay 10 to 100 times more per request.