RPA deals with structured data. AI is used to gather insights from semi-structured and unstructured data in text, scanned documents, webpages, and PDFs. AI brings value by processing and converting the data to a structured form for RPA to understand. An AI CoE is a dedicated business unit tasked with identifying AI use cases, implementing a company-wide AI vision, and deploying AI workloads on the most appropriate combination of computing hardware and software.
UIPath AI Center
UiPath AI Center is a service that allows you to deploy, manage, and continuously improve Machine Learning models and consume them within RPA workflows in Studio. This chapter deals with the subject of model deployment and management which is done through the AI Center web application available through your Automation Cloud Portal.
Where do RPA and AI meet?
• RPA has already proven its value by helping customers automate rule-based and repetitive tasks throughout an organization.
• However, a significant part of the work cannot be easily tackled by traditional RPA. We’re talking about sophisticated work involving cognitive tasks, such as classifying emails and predicting sales. This is where AI comes into play and takes RPA to the next level.
• When conceptualizing RPA and AI, it can be helpful to think of AI as the brain, and RPA as the hands. A disconnected brain can dream up concepts, but without hands, it’s impossible to apply them. At the same time, it’s impossible to handle advanced processes without a brain. It’s when the two are combined that complex tasks can be accomplished.
• This is the place for AI Center where you can automate more by adding AI to your automation.
AI Terminologies
Artificial Intelligence– The theory and development of computer systems that are able to perform tasks that normally require human intelligence and decision making.
Machine Learning– A sub-field of artificial intelligence that enables systems to learn from data.
Systems learn from previous experience and information to deduce and predict future information. To do this they use algorithms that learn to perform a specific task without being explicitly programmed.”
Deep Learning– An area of machine learning concerned with artificial neural networks. These are a series of algorithms that aim to recognize relationships in a set of data through a process that mimics biological neural networks.
Natural Language Processing-A branch of artificial intelligence that deals with analyzing, understanding, and generating human natural languages. For example, NLP enables computers to hear speech, read text, interpret the text/speech or measure the sentiment.
Computer Vision– A field of artificial intelligence that enables computers to gain high-level understanding from digital images or videos. If AI is the brain, then computer vision is the eye that enables the computer to observe and understand. It works the same as the human eye. For example, UiPath AI Computer Vision can be used instead of selectors in UiPath Studio or when we don’t have access to selectors, as it can see every onscreen element.
What is AI Center?
AI Center is an application that enables deploying, consuming, managing, and improving machine learning models. It can manage models built by in-house data scientists, by UiPath and our partners, and even open-source models.
AI Center makes it extremely simple to use machine learning in the RPA workflows built with UiPath Studio. This way, robots can process unstructured data, better handle uncertainty in decision making, and work with use cases that have tons of variables.
How does it work?
Types of machine learning models on AI Center
• Bring Your Own Model: models built by your data science team.
• Open-source model: models built by the data scientist community. Customers will be able to manage them in AI Center and train and deploy them directly in the RPA workflows.
• bullet
• Pick a model: models built by UiPath technology partners.
• Out-of-the-box models: pre-built models supported by UiPath.
Who uses AI Center?
The main persons who should have access and make use of AI Center are: Data Scientist– Building and uploading the ML models to AI Center.
Process Controller– Improving the ROI of automations by deploying models already uploaded by Data Scientists or provided by UiPath into ML skills.
RPA Developer– Consuming the available ML skills within customized RPA workflows where decisions are made by the robots.
Example
Let’s look at the context
A motor insurance company has a dedicated “Motor Insurance folder” to receive vehicle insurance claims from policyholders. The vehicles are bikes, cars, and buses. The priority of the company is the speed of providing insurance to each claim.
The motor insurance company decides to automate the process of classifying emails and placing them into dedicated folders, from where specialized teams can focus their effort on solving them. This is what the process will look like.
AI Center Main Concepts
1.Projects – A Project is an isolated group of resources, consisting of Datasets, ML Packages, Pipelines, ML Skills, and ML Logs. You may use these resources to enable building a specific ML solution for different business automation.
2. Datasets – A dataset is a folder of storage, which can have arbitrary sub-folders and files. The purpose of having a dataset is to allow machine learning models in our project to access new data points (either new files or folders uploaded from the application, or data from UiPath Robots at runtime). The Datasets page displays all the datasets within a project, along with their name, description, and creation time.
3.Data Labeling – Data Labeling is a tab where you can deploy labeling sessions to prepare the datasets for training and evaluation. Within the current version, you can deploy Document Manager sessions to build Document Understanding models.
4. ML Packages– An ML Package is a folder with all the code and metadata needed to train and serve a machine learning model. It can have multiple versions and is in some way analogous to a Package in UiPath RPA platform. Each version can have an associated change log.
5. Pipelines– A pipeline is a description of an ML workflow, including all of the functions in the workflow and the order of execution of these functions. The Pipeline includes the definition of the inputs required to run the pipeline and outputs to get from this pipeline. A Pipeline Run is an execution of a pipeline based on code provided by the user. Once completed a Pipeline Run will have associated outputs and logs.
6. ML Skills– An ML Skill is a live deployment of an ML Package. It can be used in an RPA workflow simply by dragging and dropping an ML Skill Activity in UiPath Studio. 7. ML Logs- ML logs are a consolidated view of all events related to a project.