How AI Augmented Process (A2P) will help organisations to leverage untapped potential?
AI is dawning as the new IT revolution, bringing with it the potential to change the ways of working for many, and the redefinition of business practices. Reflecting on AI projects we have led in recent years, and the results achieved by organisations who have embraced this new technology, we have no doubt of the potential economic and social opportunities it brings. Organisations need to choose if they want to be early adopters, along with the associated risks and uncertain ROI, or wait to join the late majority, risking the loss of their competitive advantage and market share. At onepoint we believe that AI is an additional lever within a continuum of efficiency solutions to meet unrealised opportunities. At onepoint we believe that AI presents unrealised opportunities to enhance existing and future efficiency solutions to meet unrealised opportunities. Further, AI presents an opportunity for human beings to move away from repetitive tasks and focus on human interactions, emotional and meaningful tasks, increasing the appeal of work for some employees and enhancing value provided to end customers. The business impacts of AI can be managed with a combination of awareness and education, as well existing tools already present in the organisation’s IT environment (see our article on this topic).
Without delay, let’s examine solutions to improve process efficiency, including the emerging option of an AI Augmented Process (A2P).
Go beyond current technology limitations through the use of A2P
When aiming to improve the efficiency of a process, an organisation has multiple levers it can utilise.
It goes without saying that the main pre-requisite to process improvement is to identify and qualify the targeted process(es). It seems easy but this step is often neglected. Discovering the value is the key starting point of a successful project, before launching any technical works (see our “Value Discovery Offering” on this topic). Once the target process(es) have been identified and qualified, the next challenge is to fine the most appropriate efficiency improvement solution: the below list not only enables better decision-making but also promises maximised efficiency and value delivery.
- Streamlining is a non-technology process optimisation method that looks to remove waste time from processes. Streamlining is the first tool of the efficiency solutions palette. By using methodologies such as Lean Thinking, organisations are able to reduce waste from its work processes without utilisation of additional technology.
- Lean thinking methodology helps to identify task duplication, unnecessary waiting time and much more. A typical example is the duplication of documentation processing in multiple systems and/or by different teams which increase time and rouse was, as well as risks associated with processes errors (manual synchronisation, knowledge not documented which could be at risk during handover, …). One approach to reducing duplication of work is to define a single source of truth for information, with processes aligned to all consume or update data in the same repository.
- Automation is a process optimisation by mechanising repetitive tasks. When the volume of processing becomes too large, Automation is the solution. Some tasks are manual, repetitive and don’t require high order decision making. Examples of tasks suited to automation could include, filling in a form with known and structured information (e.g. Name, Phone Number, Reference number, …), or sending a marketing email to a pre-defined contact list. In short, these tasks have no added value when performed by a human. Automation makes these tasks able to be performed quicker and with less chance for errors.
- Workflows are low-code/no-code process optimisation software which use automation and remove data silos. Rarely do tasks exists by themselves, without being related to other process as part of some workflow, which is where this automation opportunities presents itself. Workflow presents a user-facing interface that interacts with users and triggers automation (see our article on this topic). A simple example of this would be to extract a pre-defined client list within a CRM tool, then open an Email Marketing Software to send a prepared campaign to these clients, all with only two human clicks: the first to start the process -even optional if it is a recurring activity- and the second to validate the information before sending it.
- AI Augmented Process (A2P) are optimisation models that can process structured and unstructured data and generate pre-defined outputs (items and actions). More than processing unstructured data (search, extract, synthesise, …), AI can also generate items such as text and images. With AI Augmented Processes, employees benefit from a personal assistant doing the low value-add tasks that still need a human validation on key steps. To avoid potential issues with AI outputs being used un-checked, appropriate governance structures and ethical guidelines must be defined to ensure reasonable usage of this technology. For instance, an insurance claim operator could use A2P to suggest a classification for a claim against a pre-defined injury category list. Rather than a human action, it is the AI model that will “read” and identify the keywords of the claims.
| Technology Readiness | Business Readiness | Potential ROI | Value Assistance | Durability / Re-usability |
Streamlining | +++ | +++ | + | + | +++ |
Automation | ++ | + | ++ | ++ | + |
Workflow | +++ | ++ | ++ | ++ | +++ |
AI Augmented | + | + | +++ | +++ | + |
All of these process automation options can stand independently of each other or be used in a complimentary manner to create a tailored solution for process pain points. An appropriate Education and Awareness project should also be considered alongside any automation, workflow, or A2P change to ensure employees can realise benefits quickly, and in the case of AI Augmentation, use the solution in an appropriate manner.
If an organisation identifies that a AI Augmented Process is a suitable candidate for increasing a process’s efficiency, they next need to understand how it works and what to expect: which we will now dive into.
Deploy A2P at its full potential and with ethical safeguards
An AI augmented process’s goal is to assist employees through process-tailored AI software. As such, before starting to think about implementing A2P, the process needs to be meticulously defined. Once deployed, the AI Augmented Process will operate as described below:
- First, the A2P starts either by being triggered automatically or by a human action.
- Then, one or several tailored AIs analyse and reason about their inputs in order to create a Plan of Action (POA).
- Finally, as the A2P never takes action on its own, the POA is submitted to a human whose responsibility is to review it and either accept it or amend it.
In parallel, defining ethical guidelines is essential to ensure a fair usage of this technology which can sometimes produce results that can’t be simply explained. These guidelines must describe the auditable aspect of the technology and the process, as well at responsible ways for it to be used. Also, it should define the acceptable and unacceptable behaviours of the technology. Where possible, human feedback should be used to further improved the tailored AI.
It is also important to note that by nature, a well explained POA enables easier fixes of potential errors (less costly and time-consuming). The more transparency on inputs used, the more it will also help to improve and refine the A2P.
To conclude, the AI Augmented Process (A2P) has transformative potential for organizational reshaping and human work improvement with unparalleled power. A2P can not only streamline operations and increases efficiency but also elevates the role of human employees by freeing them from repetitive tasks. This shift allows workers to engage more in creative, strategic, and emotionally driven tasks, leading to greater job satisfaction and added value for customers. Ethical considerations and continuous learning will be key to success as organizations integrate these advanced technologies, ensuring responsible and effective use. Finally, adopting A2P as a key component of their future strategy will undoubtedly be a crucial factor for remaining competitive and innovative in an increasingly automated environment focused on human interactions.
