The rapid advancement of artificial intelligence (AI) has significantly increased the demand for AI services across various domains. However, developing AI services remains a challenging task for non-technical users, as traditional AI development methods require complex programming and data management. To address this issue, AIAP (Artificial Intelligence Automation Platform) was developed as a no-code AI tool that integrates natural language input with visual programming. AIAP helps users refine their commands through an AI suggestion feature and constructs workflows using modular, sequential steps. The goal of AIAP is to lower the barriers to AI service development, making it more accessible to non-experts. A user study with 22 participants showed that AIAP reduces workload and improves overall usability compared to existing tools.
Fig 1. A builder page for creating AI services with an example screen for Task 1 in a user study:(a) A section for entering desired instructions. Once input is provided, an $AI suggestion$ appears to interpret the prompt (a-1). Since data definition is required, it is labeled as Recorded contents. (b) A unidirectional modular step system where user inputs are accumulated step-by-step, and the service is executed in sequence.(c) The interpreted prompt. Data connections are displayed as capsules, with the action section highlighted in bold. When data is connected, it appears as AIAP_instruction.mp3. (d) The data field, which allows for the registration of files, URLs, and databases, and connects them to the prompt. (e) The action field, which automatically identifies actions from the prompt and displays related functionalities or APIs for automatic linking. (f) A menu to switch between the Service tab and the Schedule tab. (Task 1 of the User Study)}
Β Β Β \Description{Builder page for AI service creation: a modular system with user input-based interpreted prompts, data connections, step-by-step service execution, file/URL/database registration, and automated function linking.
AIAPβs design is based on insights gained from a formative study that compared the use of natural language input and visual programming methods for developing AI services. The study involved both developers and non-developers and aimed to identify challenges users face in creating AI services. Non-developers found visual programming more challenging, especially when developing complex services, while developers were more comfortable with this method.From the study, several design goals were established:
1. Modular Input Management: Breaking down commands into individual sentences to improve readability and ease of modification.
2. Minimizing Cognitive Load: Providing a linear, intuitive interface that reduces the complexity of multidimensional layouts.
3. Preventing Data and Action Omissions: Ensuring data is well-defined and linked to avoid errors.
4. Providing Templates: Offering templates or auto-complete options to help users avoid the challenge of starting from scratch.
Fig. 2. Example of results from the visual programming session. A flow for an English learning service created using post-it notes by a
non-developer and a developer. Non-developers generally find it more challenging to configure data or organize the overall flow
compared to developers. (a) Result from ND2,Non-developers outline the service flow simply, without focusing on implementation or
structure. (b) Result from D2. Developers break it down by function and provide detailed descriptions of the data and logic.
AIAP allows users to input natural language commands, which are then modularized and transformed into sequential workflows. The key features include:
1. Prompt Input Assistance: AIAP refines user commands and provides suggestions to make instructions clearer. For example, if users input long sentences, AIAP helps break them into manageable steps.
2. Unidirectional Workflow: Commands are input and executed in a top-down, sequential manner, making the workflow more intuitive compared to traditional multidimensional interfaces.
3. Automated Data and Action Linking: AIAP automatically identifies data and actions within user commands and links them to pre-trained APIs, minimizing user errors and simplifying the development process. For instance, if a user enters a command like "Check if the image contains a person and send the result via email," AIAP handles both data processing and email functionality automatically.
Fig. 3. Prompt Input and Data Connection: (a) When a prompt is entered, AI Suggest organizes the sentence and highlights Indicate
in bold. Undefined data is displayed as list of websites . (b) Add the file user want to connect in the data field. (c) When the
undefined data is selected and connected, it is displayed as image_link.xlsx .
Fig. 4. Adding and Modifying Prompts. (a) Add prompt. (b) Delete. (c) Add new prompt
Fig. 5. A builder page for creating AI services with an example screen for Task 3 in a user study:(a) The functions and APIs required
to execute the prompt have been added. (b) When the βExecuteβ button is pressed, the task is performed as shown in b-1. In this
task, the result of analyzing the list is sent via email. When βPublishβ is clicked, as shown in b-2, it analyzes the entire workflow to
automatically generate the category, title, and description, so the user doesnβt need to input them separately.
A user study with 22 participants was conducted to evaluate the effectiveness of AIAP. Participants were tasked with performing three real-world tasks: (1) summarizing meeting minutes and sending them via email, (2) summarizing and translating academic papers, and (3) analyzing images to detect human presence. Each task was performed using both AIAP and a baseline tool (Zapier). The study used both quantitative surveys and qualitative interviews to assess performance.The quantitative results showed that AIAP significantly reduced workload and improved usability compared to Zapier. Participants reported lower mental demand, effort, and frustration when using AIAP. In qualitative feedback, participants appreciated the intuitive interface and the modular approach, which made it easier to understand the workflow.
Fig. 6. The Wilcoxon Signed-Rank Test results for NASA-TLX showed that π΄πΌπ΄π demonstrated significantly meaningful improvements
over the Baseline across all items for Tasks 1, 2, and 3. A lower score indicates a more positive evaluation. β β β, ββ, and * represent
significance levels at 0.0001 < π < 0.001, π < 0.01, and π < 0.05, respectively.
AIAP is introduced as a no-code AI service development tool that combines natural language input with visual programming, allowing non-technical users to easily develop AI services. The study results show that AIAP significantly reduces the cognitive load and effort required to develop AI services, making AI technology more accessible to a broader audience. In the future, AIAP's capabilities can be expanded, and its integration with more complex AI systems can be explored.