My summary of UiPath before using it — UiPath is an agentic platform and I expect to be able to create agents that can do specific tasks.
How does UiPath explain itself in the first minute? “AI breakthroughs. Groundbreaking research. Leading AI researchers. All at UIPath.ai.”
How does UiPath work? The first step as part of onboarding onto UiPath is to create a workspace, from where I and colleagues can create and manage automations. I then enter the department I work in to as well as any services I use (e.g. Microsoft Teams or BambooHR) to receive recommendations about the best templates to use.
Out of the templates recommended to me, I’m particularly interested in UiPath’s ‘Autopilot’ capability, using natural language processing to describe a workflow that I want automated.
This Autopilot capability is based on robotic process automation (RPA) which uses intelligent automation technologies such as AI and Business Process Management to do routine tasks like data entry and filling in forms. UiPath’s Autopilot capability helps me build a prompt to automate a workflow — consisting of a trigger event, an action and context:
The workflow is then generated within the UiPath Studio where I can perform more actions to build out the workflow and submit it as a reusable template. In my example, I first need to connect the trigger to MS Teams and connect the action to Gmail before I can test the workflow.
Mine is just a simple example of taking a common set of steps and converting them into an automated workflow. UiPath’s value proposition is focused on enabling process automation at scale, applying agentic AI to use cases like these:
- Automating insurance claims processes — Using AI agents to automate repetitive processes like filling in an insurance claim, assessing the claim and payouts.
- Generating personalised financial plans and investment advice — AI agents can analyse large data sets (market trends, portfolio performance, etc.) to offer personalised tailored advice to banking clients.
- Optimising supply chain management — The example of AI agents analysing machine data in real time and predicting breakdowns to reduce downtime. Or AI agents suggesting optimal distribution routes and adjusting inventory levels based on inventory changes.
UiPath powers these use kinds of use cases through three main capabilities: process, task and communications mining — analysing data from business applications, employees’ desktop activities and business conversations respectively.
Main learning point: A couple of weeks ago, I explored agentic AI, and examining UiPath gave me a better initial understanding of how AI agents can enhance efficiency by automating common tasks and processes. With companies like Microsoft and Salesforce investing heavily in this space, it will be interesting to see how UiPath but also direct competitors such as Pega and Appian evolve their process automation offerings in the coming months.
Related links for further learning:
