In the world of business, time is money, and efficiency is the key to success. Our partnership with OutSystems has brought us to the forefront of using Artificial Intelligence (AI) to enhance business processes. One of the latest innovations we're excited about is the ability to create customized AI Agents using OutSystems ODC and the AI Agent Builder. AI is proving to be a game changer and can be a huge accelerator in industries like marine energy transportation where handling complex data quickly and accurately is critical. As we delve deeper into the capabilities of these agents, let’s explore how they function and why they are so impactful.
Simplifying Complex Tasks
Consider the traditional way of entering data for cargo and voyage details in the marine energy sector—dozens of fields to fill, one at a time, a process prone to errors and incredibly time-consuming. The power users of these applications - Operators - use their network of contacts to share information about the different shipments happening all over the world.
Operators receive messages through WhatsApp within their internal network. These messages, filled with industry-specific jargon and abbreviations, require operators to carefully discern which information corresponds to specific fields in an online form. After identifying the relevant details, they manually complete the form by entering the appropriate information into each field.
The forms used in the marine energy sector are complex, containing many fields that gather a broad range of critical data, including cargo specifications, vessel information, and voyage timelines. This complexity can be overwhelming for operators, as it demands high levels of accuracy and attention to detail. This manual process is time-consuming and prone to errors, which can lead to significant operational inefficiencies and decrease operator satisfaction.
To address these challenges, integrating an AI agent into the workflow can transform the efficiency and accuracy of data entry. The AI agent can be instructed to understand and interpret the industry-specific jargon and abbreviations found in the messages exchanged via WhatsApp. By training the AI model on a corpus of industry-specific communications, it can learn the context and nuances of the sector. This training involves feeding the AI historical data, which includes examples of messages and the corresponding fields they relate to. Advanced machine learning techniques, such as Natural Language Processing (NLP) and Retrieval Augmented Generation (RAG), can be employed to enable the AI to identify and extract pertinent information automatically.
Once trained, the AI agent can instantaneously analyze incoming messages and extract the necessary data to populate the forms. This not only speeds up the data entry process but also significantly reduces the likelihood of errors. Furthermore, by automating the initial data entry, operators can focus more on verifying the accuracy of the information before final submission, which enhances the overall quality of data that can also be used to train the model. The AI-driven approach not only streamlines operations but also supports business scalability by handling increased volumes of data without the need for proportionally increasing the staff.
What is an OutSystems AI Agent?
The OutSystems AI Agent, a key component of the AI agent Builder in the OutSystems Developer Cloud, facilitates the smooth integration of generative AI into organizations’ applications. The AI Agent is adept at automating a variety of processes. It can be customized and trained to extract and organize data from multiple sources, provide instant/real-time insights, and enhance decision-making across different business contexts.
Utilizing advanced machine learning models and natural language processing, the AI Agent boosts the functionality and efficiency of applications on the OutSystems platform, offering a more dynamic and responsive user experience.
We used the Azure AI Search with the AI Agent Builder. We’ve deployed the GPT-35-Turbo-16k AI Model available in Microsoft Azure.
After deploying the model in Azure, we configured it under the AI models in AI Agent Builder. We didn’t use the one available out-of-the-box because we wanted to control the Tokens per Minute Rate Limit.
With the AI Model configured we went and started to create the AI Agent that will be used in our use case. To demonstrate its potential, we instructed the AI agent to streamline complex tasks, such as data entry for cargo and voyage details in the marine energy transportation sector.
Selected the AI model created before – GPT 35 Turbo 16k
Provided the needed instructions for the model to act as we needed for this use case
As part of our instructions to the AI Model we defined it to act as an expert in the marine energy transportation sector and decipher the information in these messages.
The next and final step was to integrate this capability into an OutSystems application where we brought all the power of AI to streamline the existing process. Operators can now enter all necessary information into a single text area and with just one click, our AI Agent analyzes these messages, extracts the relevant data, and automatically populates the entire form.
This powerful innovation not only saves time but also significantly reduces the risk of errors. Importantly, the form is not saved automatically; this feature allows operators to review and adjust the pre-filled data, ensuring everything is correct before they decide to save the record to the database. By enhancing the data entry process and minimizing manual input, our system significantly elevates the efficiency and reliability of operations in the marine energy sector, marking a substantial step forward in maritime management.
The Business Value of AI-Driven Automation
By automating data entry processes with AI, companies can achieve several significant benefits:
Reduced Operational Costs: Less time spent on data entry means lower labor costs and freed-up resources for other critical tasks.
Increased Accuracy: Automated systems reduce human error, leading to more reliable data and fewer costly mistakes.
Enhanced Speed of Service: Faster data processing means quicker turnaround times, which is crucial for staying competitive in fast-paced industries.
Improved Employee Satisfaction: Removing tedious tasks from employees' workloads can lead to higher job satisfaction and lower turnover.
Lessons learned
As we've integrated this technology, several lessons have stood out:
Iterative Development: Constant testing and refinement are crucial. Each test teaches us something new that helps improve the AI’s performance.
Focus on User Experience: The success of technology implementations often hinges on how easy they are for end-users. The simple "Auto-Fill" button powered by AI is a hit because it makes users' jobs easier.
Broad Potential: While we started with the marine energy transportation sector, this technology has clear applications across various fields like finance, insurance, and other energy sectors, demonstrating its versatility.
Our partnership with OutSystems and our use of their AI Agent Builder application is not just about keeping up with trends; it's about setting a new standard for operational efficiency. The ability to streamline complex data entry tasks with AI is a transformative development for our clients. It signifies a shift towards more strategic business practices where AI handles routine tasks, and humans make the decisions that matter.
Comments