In an era marked by technological evolution, we delve into a conversation with James Keyworth, Director of Product Solutions at fu3e., to explore how recent advancements in AI are redefining the real estate technology landscape. Discover the core modules of fu3e.'s innovative platform and learn how AI integration enhances efficiency, accuracy and user experience, whilst also addressing critical concerns like misinformation and data security.
Can you talk about the current state of AI within the industry, given the recent advancements
Absolutely. AI has become exceptionally prevalent in the industry and within the media. The recent advancements in foundational models provided by OpenAI, Google and Meta (to name a few) have significantly changed the landscape. These models have revolutionised what we can achieve within our software and how swiftly we can introduce new features and products to the market.
Could you explain the core modules of your software and how users utilise them?
Certainly, at the core of our software, we have several key modules. Our users utilise forms and workflows to log data related to their real estate assets. This data covers aspects like cost, program, risk and other business requirements. This information then feeds into a hierarchical dashboard structure. Larger corporate clients might have a portfolio-level dashboard, which connects to fund or sector-level dashboards and, eventually project-level dashboards.
How does AI enhance your software's capabilities and benefit your users?
AI contributes to our software's capability to make users more productive and automate our solutions. For instance, one upcoming feature involves integrating with a large language model. This integration enables users to prompt the language model, similar to interacting with ChatGPT. This assists them in drafting reports efficiently. For example, when creating a project execution plan, they can provide data, ask questions, and receive innovative ideas from the language model, enhancing their efficiency and accuracy.
What technology are you using for this integration?
We're integrating with the OpenAI API.
Given the concerns about misinformation and inaccuracies generated by AI models, how do you plan to address potential challenges in this area?
Addressing misinformation is a significant ethical concern with AI. To ensure accuracy and reliability, we emphasise maintaining a human element in the solution. While AI is a tool to enhance user productivity, any information provided by the AI must be reviewed and analysed by the document's author. Transparency about the limitations and capabilities of AI models is crucial to avoid generating inaccurate content.
Is there a risk of AI-generated misinformation, and how are you planning to manage it?
Yes, the risk of AI-generated misinformation, also known as hallucinations, is a concern. It's important for any software working with these models, including ours, to provide sufficient information to users about the limitations and capabilities of AI. Educating users and promoting responsible use of AI-generated content is a critical part of our approach.
How far along are you in implementing these AI features into your software?
We're in the advanced stages of developing the first AI feature, and it's expected to be available within the next month or two. We’re also exploring other ideas and concepts related to AI within our product roadmap. AI will be a significant focus in our product strategy for the upcoming year.
Can you discuss other concepts or features you're considering apart from the current AI integration?
Our focus is on where we can provide the most value to our customers. One prominent trend is the automation of day-to-day tasks. Executives often have limited time and benefit from features that enhance productivity. For instance, we're exploring summarising uploaded documents or reports for them. Additionally, personalisation is on our radar, tailoring the user experience based on interactions with the software.
Have you received direct feedback from customers expressing interest in AI, or is this more driven by your internal innovation efforts?
While we haven't had explicit requests for AI features, we regularly receive feedback from customers seeking automation and efficiency enhancements. Our approach is to analyse these requests and determine how AI technology can fulfil those needs. It aligns with our broader product strategy of increasing productivity, data insights, and connectivity.
Considering the potential scepticism around AI, do you plan to provide training and education to users as you roll out these features?
Providing effective communication with the AI models, such as prompt engineering, is essential for users to achieve optimal results. We intend to guide and assist users in effectively communicating with the AI for features where this interaction is crucial.
Could you clarify the data usage and security aspects of integrating with OpenAI's API?
OpenAI's API operates differently from models like ChatGPT. The data provided to the API is not used to train the model, ensuring that user data remains secure and private. The data policy of OpenAI is user-friendly and respects data security.
What is the overarching goal of integrating AI into your software?
Our goal is aligned with our broader product strategy: to provide a platform that enhances user productivity, delivers data insights, fosters connectivity, and ensures security through digitised processes and reporting. Integrating AI supports this strategy by offering features that align with these objectives.