When it comes to adopting AI in business, we believe there is a useful distinction to be made between how AI can help individual employees and how it can be used in the back office to intelligently automate business processes.
There are different methods, tools and considerations for these two types of AI deployment, and thinking about your business's AI strategy through these two distinct lenses can make the opportunities for AI adoption much clearer.
Individual Automation
When most people think of AI, they imagine a ChatGPT style experience. This usually involves a conversational UI used by a single employee to ask questions, generate content, or automate a piece of work.
In this mode, AI serves as a productivity tool that can make individual employees and knowledge workers more efficient and effective in their work.
The task for business leaders is to think about how they can help their employees take this further. Which tasks can AI best assist with? Which specialised UIs do we need to build for them? Which data sources do we need to connect to? How do we help our employees make use of AI without exposing us to risk?
To be clear, we don't think that individual automation is as simple as just giving your employees access to ChatGPT or Claude in an unconstrained way. A much more measured, thoughtful and integrated approach is required in order to harness the full value of AI at an individual level, and most businesses are still at the first level of maturity on this journey.
Business Automation
An area that has received less attention is in using AI to automate business processes. Here, AI operates “in the back office,” processing requests and handling workflows and tasks that would have previously required human effort.
In this mode, AI is in the driving seat, carrying out work on our behalf and potentially only involving a human for approval or when a situation of interest is detected.
This type of AI could be deployed as a fairly standard piece of automation, or perhaps via an autonomous agent which is mimicking the behaviour of an employee.
With this type of automation, there is no employee accessing a user interface, and indeed this type of automation may not need a user interface at all.
In this world, API connections into your business systems are likely to be more relevant as we don't have the human translating data for us. Guardrails and monitoring are also likely to be much more important, as humans will be much less involved in the task.
We think that the issue holding back businesses is a lack of understanding and case studies in how AI can be deployed in this mode. Most businesses are so focussed on Chatbots and individual automation, that they are not uet looking at the potential of using AI for automation and passive monitoring of your business data. We think this will change in the next few years as enterprises look to do more with AI and as autonomous agents grow in adoption.
Team Level Automation
Another interesting perspective is considering how AI can help teams perform better as a unit. Although there has been less work in this area, it seems an obvious next step.
Firstly, our team-based AI tools might look like individual automation tools but be much more collaborative, similar to tools such as Google Docs or Figma. If team members are working on a task using AI, having it confined to one person’s account behind a single-user UI is a step backwards from todays multi-player UIs.
We also expect that prompts will eventually shared across teams and up to the enterprise level. For instance, if our business has an AI-powered onboarding process, it’s crucial that this process is complete, vetted, and consistently followed across the organization, rather than individuals creating their own ad-hoc prompts.
Finally, one promising use case we've identified is in using AI for intelligent prioritsation fo work. We have proposed many UIs with "intelligent in-trays" filled with tasks prioritised such that the next team member to pick up and work on based on the importance of the task.
Though these are early ideas, the concept of a team runs deep in todays organisations, and we think that there will be a set of team level use cases that sit between the individual and the back office.
Work Will Flow Between All Three Modes
With these three types of AI deployed, we can see a scenario where all three combine to accelerate end-to-end business processes.
Consider an end to end business process associated with onboarding a new employee:
- Individual automation - An HR team member talks to the new hire over the phone to capture demographic details, work history, and a skills profile. The voice call is transcribed by an AI model and converted into structured data. The HR member can then review the transcription and submit the record into the HR system through their copilot.
- Business automation - An AI agent identifies the new employee record and automatically sends reference requests to previous employers by email.
- Business automation - A seperate agent looks at the person’s job role and skills profile and automatically enrolls them in relevant training courses.
- Team automation - The training team could use a collaborative AI powered interface to personalize the training schedule for the employee. They make use of their prompt library to ensure a consistent approach to upskilling across the global business.
- Individual automation - The employee could finally use some individual automation to complete their personalised training, asking the AI to prompt them with quizzes and generate example sceanarios to help them pass their upcoming exams.
This is a hypothetical example, but it shows how each step of the process is accelerated by intelligent automation whether the work is currently sitting with the individual, team or in the back-office.
Use This To Structure Your AI Strategy
We have thought a lot about how businesses will adopt AI, and we find this distinction between individual, team and business level automation to be a really useful one. Please get in touch with us to learn about how we use this framework to help you uncover use cases in your own business.