For decades, business leaders were told that the latest technologies available would revolutionise their operations, and those that lagged in technology adoption would suffer as a result. While there is a great deal of truth to these sentiments, the reality has proven to be far more nuanced. Despite the lessons learned by early adopters of the “must-have” technologies of the past few years, it’s become standard operating procedure for most companies to immediately look at investing in the most recent tools available, perpetuating the familiar hype cycle.
AI, for example, has become the latest business buzzword. Organisations of all sizes and from all sectors are adding AI to their technology portfolios, looking for immediate results. Unfortunately, all AI solutions are not created equal, and many companies are struggling to make their investments deliver the promised outcomes.
While there’s no doubt that delaying AI adoption can put a business at a competitive disadvantage, achieving success requires more than just investing in an AI solution. Success also rests on integrating legacy and modern technologies, risk management and access to the right skills. Another vital factor to consider when it comes to AI implementations is addressing ethics appropriate for the industry or ecosystem.
In other words, companies need tailored, fit-for-purpose AI solutions in order to achieve results.
The right tool for the job
Generative AI (GenAI) like ChatGPT has helped AI become a mainstream technology, but companies are starting to find that GenAI solutions are limited on their own. Generative AI in action would be an HR service chatbot that can answer common questions and provide information to employees/users based on their input. A compound AI combines the capabilities of multiple agents to provide more comprehensive assistance to the user.
As an example, let's consider an employee named Nicole who wants to go on leave and needs to know the weather patterns for her planned travel, as well as how much leave she has available.
In this scenario, the compound agent seamlessly integrates two separate agents – a weather information agent and a human resource management agent – to provide Nicole with the information she needs.
First, Nicole interacts with the compound agent and informs it of her intention to take a holiday. The compound agent then utilises the weather information agent to pull up the forecast for the location of her holiday, providing Nicole with the necessary information to plan her trip accordingly.
Next, the compound agent accesses the human resource management agent to retrieve information about Nicole's current leave balance. It informs her of how much leave she has available and helps Nicole schedule the holiday accordingly.
By combining the capabilities of the weather information agent and the human resource management agent, the compound agent provides Nicole with a more comprehensive assistance, helping her plan her time off more efficiently and effectively. Overall, a compound agent enhances the user experience by seamlessly integrating the capabilities of multiple agents to provide more comprehensive assistance to the user.
Agentic AI is a compound agent that essentially integrates with databases and external tools to enhance problem-solving capabilities and adaptability to scenarios that would typically come in those use cases. Designed to handle incredibly complex and repetitive tasks across various business functions, agentic AI solutions like iOCO’s AI agents can analyse huge volumes of data, understand relationships, provide visibility into operations and support better decision-making. Use cases include fraud detection, customer service, supply chain management, compliance and risk management, and logistics optimisation, to name a few.
It's all about the data
Data is at the heart of a successful AI solution. Data improves decision making, can be used to automate tasks, and can provide hyper personalised customer experiences, but many companies continue to struggle under the weight of traditional business models and analogue business processes, inhibiting the potential of data analytics and AI. Some have started modernising, but can't make the cultural shift that's required – or commit to the information management, advanced skills and technology investments that are needed.
Considering that each AI use case has its own data requirements, companies must invest in their data practices in order to gain the benefits of any AI investments. Business and technology leaders should analyse industry-specific examples of how data and analytics can provide economic benefits for the organisation, and measure the value of the organisation's data assets to help shift the company’s mindset to data as a corporate asset. Decision-makers in charge of the data practice should insist on being involved in corporate and strategic planning so that they can make sure that data is at the forefront of how they execute those plans. This should be communicated internally and publicly through annual reports or investor analysis and conferences.
Ideally, data and analytics strategies should be a regular and routine discussion topic in the boardroom, in every industry, so that organisations can use data analytics as competitive weapons, operational accelerants and innovation catalysts. This is going to become even more important as AI continues to filter through all aspects of corporate operations.
Ethical considerations
No conversation around AI is complete without a discussion around ethics. As organisations start to collect and use more data, they need to be mindful of the ethical considerations, security, appropriate use and AI hallucinations or biases in the algorithm. AI hallucinations occur when an AI system generates incorrect, misleading or nonsensical information in response to a user prompt.
Data, AI and digital technologies must therefore be used in a responsible and ethical way – not only to avoid embarrassments, but to build trust with stakeholders and customers. A good way to do this is to publish an ethical framework which includes privacy, discrimination, bias and accountability. As AI tools get smarter, it’s going to get easier to forget that we’re talking to an AI agent, so companies must be transparent and explicit about how the technology is being used.
Organisations that have started on their AI journeys have already discovered that there are a multitude of factors that need to be taken into consideration to ensure they get the most out of the technology. The keys to success, however, rest on the data the system uses and the type of AI solution implemented. Once companies get these two vital aspects right, AI can transform the business by providing actionable insights from vast datasets, automating decision-making processes, personalising customer experiences, generating creative content and enabling innovative solutions that drive growth, efficiency and competitive advantage.
Share