Amplifying Business Process Automations with UiPath and Amazon SageMaker

By Scott Schoenberger, Sr. Product Manager – UiPath
By Daya Thakur, Sr. Partner Solutions Architect – AWS
By Jagjit Dhaliwal, Principal Partner Development Manager – AWS

UiPath

As organizations continue to embrace digital transformation, they are increasingly turning to intelligent automation technologies to streamline their business processes and improve efficiency.

UiPath is an AWS Partner and AWS Marketplace Seller that’s a leading provider of enterprise automation solutions, and Amazon SageMaker is a fully-managed service that provides developers and data scientists with the ability to build, train, and deploy machine learning (ML) models quickly.

In this post, we’ll explore how UiPath Business Automation Platform (UiPath Platform) and Amazon SageMaker can be integrated to help businesses automate complex processes, improve decision making, and drive innovation by leveraging the power of artificial intelligence (AI).

About UiPath

UiPath holds the Amazon SageMaker Ready specialization and offers an end-to-end platform for automation with enterprise-ready cloud infrastructure, AI services, and intelligent automation solutions on Amazon Web Services (AWS) that provide the foundation to scale automation.

This can help accelerate deployment, integrate with AWS AI services, and leverage built-in intelligent automation solutions such as customer experience in the contact center and document processing.

UiPath Platform on AWS can be quickly deployed using deployment accelerators:

Solution Overview

The UiPath solution allows customers to bring machine learning inference from Amazon SageMaker directly into their business automation to simplify automation deployment pipelines and operationalize ML insights.

Data science teams using SageMaker to build, train, and deploy ML models can connect to UiPath Platform to quickly and seamlessly connect new models into business processes, without the need for complex coding and manual effort. The joint solution will help customers rapidly deploy ML models into production, optimize productivity of data science teams, and increase the speed of innovation.

UiPath Platform and SageMaker integration offers several benefits to businesses:

Enables organizations to automate complex processes that were previously difficult or impossible to automate. By leveraging machine learning algorithms, businesses can automate decision-making processes that require human intelligence, such as fraud detection, insurance policy underwriting, and anomaly detection.
Integrating UiPath Platform and SageMaker improves the accuracy of automation workflows. By using ML models to analyze data, businesses can make more informed decisions and reduce the risk of errors.
Allows businesses to scale their automation efforts, as data scientists can train ML models on millions of data points from multiple systems using Amazon SageMaker tools and then apply the inference from ML models to various automation flows within UiPath Platform. SageMaker provides a scalable platform for building and deploying ML models while UiPath Platform provides a scalable platform for automating business processes. By combining the two, businesses can create intelligent automation solutions that scale to meet their growing needs.

Reference Architecture

UiPath Platform connects with Amazon SageMaker APIs using the SageMaker Software Development Kit (SDK). A design-time lookup, with SageMaker’s ListEndpoints API, in UiPath’s SageMaker activity (Get Inference) lets users pick a deployed model they want to run the inference against.

Users can configure input/output through an easy-to-use configuration form on UiPath Platform. At runtime, SageMaker’s real-time endpoint is invoked through InvokeEndpoint API for the selected model, and results can be passed to a downstream business process or application for actions based on the result.

Figure 1 – UiPath Platform and Amazon SageMaker integration.

Real-time inference is ideal for inference workloads with interactive and low-latency requirements. Users can deploy ML models on SageMaker hosting services and get an endpoint that can be used for inference. These endpoints are fully managed and support autoscaling.

Using SageMaker, data scientists have the flexibility to choose between a low-code or high-code approach when building ML models. For a streamlined process with minimal coding, they can utilize SageMaker Studio with Autopilot, enabling them to develop and train ML models efficiently. Alternatively, data scientists can opt to write custom code within SageMaker Notebooks to construct and train models for their specific requirements.

SageMaker’s no-code tools, like Canvas, empower business analysts to build ML models. They can collaborate with their data scientists and MLOps teams to deploy these models in production. Moreover, by utilizing the “Get Inference” activity in UiPath, business analysts can seamlessly integrate these models into their automation workflows.

UiPath’s Integration Service offers Robotic Process Automation (RPA) developers and citizen developers alike the ability to authenticate, govern, and share connections to a wide variety of enterprise software-as-a-service (SaaS) applications across their teams.

Offering curated activities and event triggers that don’t require deep understanding of vendor APIs, Integration Service takes the guesswork out of API-based automations. Furthermore, Integration Service allows users to build their own connectors to services that aren’t yet in the catalog, combine API activities with traditional user interface (UI) activities, and incorporate other innovative AI activities from Document Understanding, Communications Mining, and AiCenter.

A connection to your AWS account from UiPath Platform needs to be established before using Amazon SageMaker. The connection can be established in your Automation Cloud/Integration Service tenant or within Studio Desktop and Studio Web at design time. There are no limits to the number of connections, and they can easily be shared across folders to govern access and limit sharing of sensitive credentials. In this way, connections can be established to development, staging, and production environments for requisite testing.

Use Cases for UiPath Platform and SageMaker Integration

There are various use cases for integrating UiPath Platform and Amazon SageMaker. Here are few examples:

Generative AI applications: Automation developers, including citizen developers, can include generative AI capabilities including text generation, text summarization, and question-answering using foundation models deployed with SageMaker Jumpstart. For instance, developers can build an automation workflow using UiPath Platform to read the emails and provide a summary to the end users.
Fraud detection: By integrating UiPath Platform and SageMaker, businesses can create an intelligent automation solution that detects fraudulent transactions. ML algorithms can analyze transaction data to identify patterns that are indicative of fraud, while UiPath Platform can automate the process of flagging suspicious transactions for review.
Insurance policy underwriting: UiPath robot sends policy application data in real-time to a SageMaker-hosted insurance risk ML model which returns a policy risk calculation score informing the business user whether the policy can be approved.
Predictive maintenance: ML algorithms can analyze data from sensors and other sources to identify patterns that are indicative of impending failure, while UiPath Platform can automate the process of scheduling maintenance before the equipment fails.
Anomaly detection: ML algorithms can analyze data to identify patterns that are outside the norm, while UiPath Platform can automate the process of investigating the cause of the anomaly and taking corrective action.

Example: Fraud Detection Workflow

More and more businesses are moving online; at the same time, fraud is increasing as well. Fraud causes substantial annual losses amounting to billions of dollars. However, with the aid of the ML-based solution built with UiPath Platform and Amazon SageMaker, it’s possible to automate detection of fraudulent activities from vast volumes of transactions.

UiPath robots can collect data from variety of enterprise applications and upload to UiPath data service storage. The transactions are then sent to SageMaker for identifying fraudulent activity. Analysts only have to review the transactions that have a high probability of fraud. This reduces manual effort and improves the quality of review.

The integration between UiPath Platform and SageMaker also reduces the time required for developing or implementing this solution. Customers can train their own custom model or use pre-built fraud ML models available on SageMaker JumpStart to further accelerate development cycles.

Figure 2 – Fraud detection workflow with UiPath Platform and SageMaker.

Users can follow the steps below to operationalize their machine learning models, deployed through SageMaker, on UiPath Platform:

Define your use case and associated business process.
Deploy UiPath Platform on AWS using the UiPath partner solution template to deploy the UiPath Automation Suite platform.
Identify data sources for your ML problem and train a custom model using SageMaker, or choose a pre-trained model from SageMaker JumpStart.
Deploy ML model to SageMaker real-time endpoint for inference.
Configure SageMaker Activity package on UiPath to establish authentication between UiPath and your AWS account.
Configure “Get Inference” activity to point to the deployed model.
At runtime, invoke “Get Inference” activity in UiPath workflow and pass the inference for downstream business processing.

Figure 3 – “Get Inference” activity configuration form on UiPath Platform.

Conclusion

Integrating UiPath Platform and Amazon SageMaker enables businesses to create intelligent automation solutions that streamline business processes, improve decision-making, and reduce the risk of errors. By leveraging the power of machine learning, organizations can automate complex processes that were previously difficult or impossible to automate.

This post covers the reference architecture for UiPath Platform and Amazon SageMaker integration, example use cases, and steps to operationalize ML models using this solution.

To get started with UiPath on AWS, use AWS Quick Starts or AWS Marketplace.

Resources below can help you start your automation journey with UiPath and AWS:

.

.

UiPath – AWS Partner Spotlight

UiPath is an AWS Partner and leading provider of enterprise automation solutions. UiPath is leading the automation first era—championing a robot for every person and enabling robots to learn new skills through artificial intelligence.

Contact UiPath | Partner Overview | AWS Marketplace