Best Practices and Best Uses of Chat Packages

Interacting with AI-powered chat systems has gained immense popularity in recent years. Businesses across various industries are increasingly adopting these technologies to enhance customer engagement and streamline access to critical information. By setting up AI-driven chat capabilities, companies enable their customers to interact with essential resources such as product documentation, user manuals, research articles, and other business-critical content. This approach not only improves customer experience but also empowers users to find answers and solutions quickly and efficiently without the need for direct human intervention.

Additionally, these chat systems offer businesses the opportunity to provide round-the-clock support, reduce operational costs, and gather valuable insights into customer behavior and needs. As a result, AI-powered chat systems are becoming a pivotal tool in modern customer service strategies.

But it is no simple undertaking to set up chat for your customers. AiWerkz makes it easy and inexpensive. You have customers and documentation. Your customers want to chat with your documentation. And that is what AiWerkz does.

Quick overview on what AiWerkz does

Lots of people are familiar with Ai chat systems like ChatGPT. We found that people often paste content from a web page or a PDF into an AI chat system like ChatGPT. And that gave us the idea that we could improve that process by packaging content in a format that is both easily accessible and works well with AI chat systems. And that is how we invented AI chat packages.

A chat package is a thing that you can copy and paste into any AI chat system. The chat package has data in it of three different types: 1) instructions to the AI chat system, 2) content to chat about, and 3) a question from the user.

Using a chat package is super simple. The user clicks on a button or a link and the chat package is copied into the user's clipboard. Then the user pastes the chat package in the AI chat system of their choice. Then chat away. As easy as that. No fancy or complicated technology for the user or the creator of the chat packages.

Creating chat packages

Creating a chat package is super simple. Click the menu then select from one of several different ways to create a chat package. See How to Create an AI Chat Package

There are really two different types of chat packages.

• Saved chat package where you create a chat package with content. This type of chat package has a unique ID and URL that can be accessed by your users.
• Ad hoc chat package where a chat package is created based on the content of the web page it is attached to.

The main thing to remember is that a chat package is how information or content is organized and made accessible.

NOT the Best fit for AiWerkz chat packages

A chat package is NOT best for use cases where websites need to keep a copy of chat sessions with users. For example, if a hosting organization needs to retain a copy of the chat session, then that does not work with chat packages. The chat session between a user and a chat system using a chat package is private. So, chat packages are not any good if your organization hopes to harvest personal data from chats.

A chat package may NOT be the best solution if strict controls are required to keep the conversation rigidly bound to specific content. The nature of chat packages is that they may drift or swerve away from the topic at hand. We provide instructions to the chat system to minimize that, but it is a risk. AI chat systems can be unpredictable.

Best fit for AiWerkz chat packages

Chat packages are great if you have web accessible content that your user's need to chat with. The content could be PDFs, static web pages or dynamically created web pages.

Long form content, like greater than 1000 words, that your users have to read through to get their questions answered. With a chat package, the user just enters their question and then they get an answer that they can chat with.

Many small pieces of content. In some cases, content may be spread across many web pages. Assembling that content together in chat packages or chat package groups provides a very simple way to create a chat experience fro your users.

Automatically enable chat for every web page. AiWerkz chat packages can simply be "turned on" for every page of a web site. At the top of the web page is an AiWerkz chat header, a text box where you can enter a question, and then "chat" with this web page. This same capability can be enabled on your web site.

Customizable context for each web page. As just mentioned, the content on a web page, even if dynamically created, can be formed into a chat package. But that chat package can also include other content to provide a rich and context deep chat experience.

Did we mention that we are easy? Take a look at our learning hub or just enter a question above to learn about how to enable chat for your users.

And did we mention that we are cost effective? Compare us to other AI chat solutions. Our prices are so low because 1) we offload chat sessions to the user's own Ai chat system, 2) no complicated servers to set up, 3) no expensive subscriptions to maintain, 4) no need for AI scientists and 5) no long term commitment.

Developing a chat package strategy

A good chat package strategy will help the users of your site to more easily get answers to their questions while being easy to maintain. A chat package strategy includes organizing content, integrating chat packages into your web site, communicating with your users about the use (and benefits) of chat packages and ongoing maintenance.

A web site is typically designed to help your customers find information that they are looking for. At AiWerkz we want to leverage that existing design. Web sites will have either an organized structure that directs users through menus or some sites may also have a search engine. Both methods work well with AiWerkz chat packages.

A person interacting with a web site already has an idea, a topic, that they are searching for. And it is likely that your web site already does a good job of getting the user to the content. This means that the scope of the content that the user is interested in is somewhat defined. The objective then becomes to provide information, in the form of a chat package, based on that scope.

Suppose a user has arrived at a web page based on either a search query or clicking through a menu system. It is now known what the user is interested in. The page that they have arrived at will, hopefully, have content to answer part or all their questions. When AiWerkz chat is integrated into your web pages, then the user can enter a question, retrieve a chat package and paste it into the AI chat system of their choice. Furthermore, the AiWerkz chat interface can be configured to capture the text of the web page and include it in the chat package.

The result is that the simplest design strategy is to just have the user enter a question into the AiWerkz chat box and retrieve a chat package that includes the content of that web page. But it is very likely that there is related content that would be beneficial to include in the chat package.

Content, up to five to ten thousand words can be converted into a chat package. That chat package, identifiable by its unique document_section_id, can be accessed as a standalone chat package or it can be included along with other chat packages into another chat package.

Since chat packages can be stand alone or integrated together into a larger chat package, it is important to be aware of the overall size of the chat package especially if the chat package is made up of multiple chat packages.

Chat packages can be organized in several different ways. For example:

• Each chat package can be referenced as a URL. This means that you can provide links or lists of links on your web pages. If you are a good with JavaScript then you can make buttons where the user clicks on a chat package and the contents of the chat package are automatically copied to the user’s clipboard with an instruction like “A chat package has been copied to your clipboard, paste it into ChatGPT to chat with it.
• Chat packages can be referenced as a group and accessed through the chat interface. Several chat packages can be assembled into a group, a group_id specifically, and then access through the AiWerkz chat page. Take a look at AiWerkz Chat Page where the user can see and select the available chat packages.
• AiWerkz Chat embedded in web pages. A chat text field and chat button are included on some or all of your web pages. Meta tags will be configured to tell AiWerkz how to assemble a chat package and may include the page contents or saved chat packages.

The first step is to create several chat packages and get a feel for how they work. It may strike some as odd that AiWerkz makes use of the clipboard to get the chat paclage to the user's Ai chat system (like ChatGPT), but will likely quickly realize how efficient this method is. Try it out!

There are several options to integrate chat packages into your website. Keeping in mind that chat packages are either saved chat packages or dynamically created when AiWerkz is embedded within web pages.

For web sites that have web pages that have a lot of content in each page, integrating the AiWerkz Chat interface into your web pages might be the best option. This is because the user will have arrived at a web page with content that is likely sufficient to answer the user's question. See How to add chat to a web page

For web sites that have many pages but with little content then creating saved chat packages that are referenced from the web page with a link or button might be the best option. An example of this type of site is one that references PDFs or other downloadable content. The PDFs could be created into chat packages and then made accessible from the referring web page. An example is Open the chat package titled "AiWerkz: Revolutionizing Document Chat"

For web sites that have fewer web pages or limited navigation or more complex questions or topics then using the AiWerkz Chat Page. This works best when there may be up to 20 chat packages assembled under a group_id. The user can then select the chat packages that best align with their question. Then a chat package will be created for them.

AiWerkz chat packages may seem unconventional as they rely on users copying a pre-packaged data structure (chat package) and pasting it into AI chat systems like ChatGPT. This design prioritizes simplicity and user control. By copying and pasting, users can engage directly with the content without requiring complex setups or additional software. It ensures privacy, eliminates unnecessary infrastructure, and keeps interactions intuitive, leveraging a method most users are already familiar with.

Why Use AiWerkz Chat Packages Instead of Manual Copy-Pasting?

Manually copying content with a mouse can work for simple, short documents but becomes impractical for larger, complex, or multi-section content. AiWerkz chat packages streamline this process by:

• Structured Data: Packages include not just content but instructions and user questions, ensuring AI systems respond contextually.
• Efficiency: Automates assembling long or multi-part documents, saving users from manually piecing sections together.
• Consistency: Avoids errors or missed sections during manual copy-pasting, especially with long-form content.
• Scalability: Enables handling multiple documents, dynamic content, or grouped packages, which is tedious to manage manually.
• Ease of Use: A single click copies everything needed-content, instructions, and prompts-directly to the clipboard for immediate use.
• Specialized Instructions: AiWerkz includes and even manages instructions to the AI chat system to optimize the chat session.

This efficiency and reliability make AiWerkz ideal for repetitive or complex document interactions.

• Elimination of Complex Infrastructure: Competitors rely on Retrieval-Augmented Generation (RAG) systems, which require expensive and technically complex setups like vector databases and middleware for managing queries. AiWerkz bypasses all of this by leveraging the user's existing AI chat system (e.g., ChatGPT) and simply packaging content into a structured, portable format. No additional backend infrastructure is needed.
• Reduced Maintenance Effort: Traditional RAG systems require ongoing management, including re-indexing, vectorizing new data, and maintaining middleware. AiWerkz eliminates these tasks by using static or dynamically generated chat packages, making updates straightforward. For example, editing a document or adding new information involves modifying or generating a new chat package-no re-indexing or recalibration of systems is necessary.
• Scalability Without Added Complexity: AiWerkz chat packages are lightweight and portable, designed to scale across hundreds or thousands of documents with minimal overhead. Packages can be grouped or organized dynamically using group IDs, enabling easy categorization and access without introducing complex hierarchies or dependencies.
• Privacy and Data Control: Unlike competitors that pass queries and content through intermediaries, AiWerkz keeps data interactions between the user and their AI chat system. This reduces compliance concerns and simplifies processes for industries with strict privacy requirements (e.g., healthcare or finance).
• Cost-Effectiveness: AiWerkz offloads the AI processing cost to the user's own AI account, eliminating the need for costly AI subscriptions or server management by businesses. This keeps operational costs low while providing scalable solutions.
• Streamlined User Experience: AiWerkz's focus on using clipboard-based chat packages means users can deploy and update packages without learning complex workflows or tools. Simple JavaScript buttons or links allow seamless integration with existing websites.
• Future-Proof Design: As LLMs expand their context windows (now capable of processing over 20,000 words), AiWerkz becomes even more effective, eliminating the need for document chunking. This future-proofs the system against evolving AI technologies.

Practical Example:
Imagine managing FAQs for a product line. Using AiWerkz, you:

1. Organize content into chat packages for each FAQ.
2. Update individual packages as needed—no need to re-index a vector database.
3. Offer users instant access through simple web page buttons or links, without additional software or workflows.
4. Automatically integrate multiple chat packages into a single conversation using group IDs, making cross-references easy.

By keeping the process intuitive and infrastructure-light, AiWerkz ensures businesses focus on delivering value to users, not managing complex systems.