Session 3 - Specialized AI Associate Series: AI Powered Automation through specialized AI

🚀 Welcome to Session3 / UiPath Upskill to Agentic Automation - Specialized AI Series 2025!

In this session, we will start by deep-diving into Orchestrator functionalities: manage all automations, robots, and automation assets from the same place. Then we follow-up by seamlessly integrating API automation and UI automation within the same design environment using UiPath Integration Service.

📕 Agenda:

Introductions
AI Powered Automations: Generative vs Specialized
UiPath Specialized AI Components (High Level - AI Center, DU, CM)
Identifying Specialized AI Use Cases
End-to-end Walkthrough with AI Center
Q&A (i.e. topic or exam prep questions)
For links to the self-study modules, go here: https://academy.uipath.com/learning-plans/specialized-ai-associate-training, and click on the second tab.

1. The UiPath word mark, logos, and robots are registered trademarks owned by UiPath, Inc. and its affiliates. UiPath (R) is a registered trademark
in the United States and several countries across the globe. See TMEP 906. ©2025 UiPath. All rights reserved.
By UiPath Community
AI Associate Developer Series -
Session 3
AI Powered Automation through
specialized AI.

2. Speakers
2
Russel Alfeche
UiPath MVP, Technology
Leader, Automation @qBotica
Priyadarshini Estevez
UiPath MVP,
AI Strategy & Solutions Consultant
Naveen Chatlapalli
UiPath MVP,
Solutions Architect@ Ashling Partners

3. 3
Agenda slide
01
02
03
04
Automation and AI in UiPath: Generative and Specialized AI
Specialized AI Platform Components
AI Center Overview & Deployment
Identifying AI Center Use Cases

4. 4
Generative AI Vs Specialized AI
Generative AI Specialized AI
What is it? Type of AI that can create new
content, such as text, images, or
music, by learning patterns from
existing data.
Designed to excel at specific tasks or
domains, delivering highly accurate
results
How does UiPath leverage it? Large Language Models (LLMs), a
type of Generative AI that has been
trained on huge amounts of text, like
books, articles, and websites, to
understand and generate human-like
text
Enterprise automation, enhancing the
automation of documents, screens,
tasks, and processes.
Example A popular example of LLM is Chat
GPT, helps in answering questions,
writing essays, translating languages,
and even chatting with us.
Invoice extraction, where UiPath has
created ML models specialized in
identifying and extracting data from
invoices from multiple countries.

5. 5
AI-powered Automation in UiPath
Products
Generative AI Specialized AI
Open | Flexible | Responsible
Action
AI Trust Layer
UI
API
Human
in the Loop
Context
Documents
Data
Processes
People
Communications

6. 6
Machine learning and natural language processing concepts
Source: https://www.educba.com/nlp-and-machine-learning/

7. 7
AI-powered automation
Open | Flexible | Responsible
Supported by UiPath Built with UiPath or BYO
BYO
Docs Screens Tasks Processe
s
Solutions
AI Infrastructure
Integration Service –Validation Station –Active learning –Fine tuning –Guardrails –Auditing
AI-powered automation
Delivering enterprise automation with Specialized AI
Generative AI Specialized AI
Context
HITL
UI
API
Action
People
Comms
Docs
Data
Processe
s

8. 8
Produce a digital
copy of any
document, classify it,
and extract the
relevant data,
involving humans
when needed and
learning from the
interactions.
Platform component
used for managing,
training, deploying
and retraining ML
models.
A powerful, no-code
NLP solution that
combines
unsupervised
learning with
employee-led active
learningto create
models that fully
understand the
context of our
customers'
businesses.
Intelligent Xtraction
and Processing, is a
multi-modal data
extraction capability.
IXP converts data
from unstructured
documents and
communications to
structured data for
automation and
analytics.
The UiPath Platform components with Specialized AI

9. 9
Intelligent document processing with GPT
Read the email and identify intent
Process the Application Form
1
3
4
Business process: Loan application processing
Verify and extract data from the ID Card
Verifyand extract data from the Employment Letter5
Large language model (LLM)
Specialized AI model trained on documents pre-labelled with GPT
Specialized AI model for ID Cards
Question answering powered by GPT
Classify the attached documents2
Classification powered by GPT
Generative AI
Generative AI
Generative AI

10. 10
AI Center Overview &
Deployment

11. 11

12. 12
How AI Center work in your Solutions ?

13. 13
AI Center Core Concepts
Roles
Create/Edit
Delete
Projects
ML Packages
Datasets
Data Labelling
Training
Evaluation
Full
Pipeline
ML Skills
ML Package Validation
Datasets
Pipeline
ML Skill deployment
ML Skill prediction
ML Logs

14. 14
Automation
Cloud
Automation
Suite
On-premises
Air-gapped
On-premises
Online
Hybrid mode
(Cloud AI Center +
Orchestrator on-
premises)
Deployment and Installations
•Deploy, manage, and
improve ML models.
•Requires no
infrastructure and no
maintenance.
•Internet access is
required.
•Deploy the full UiPath
Automation Platform in the
environment of your choice,
includingon-premises
Virtual Machine
infrastructure, or cloud
subscriptions
•Automation Suite contains
everything in one package
that you can deploy in multi-
node mode.
•Deploy, manage, and
improve ML models
locally.
•Customer-managed
infrastructure.
•Fully integrated with on-
premises Orchestrator.
•All resources must be
manually downloaded and
then loaded into the node.
•No internet access is
required
•Deploy, manage,
•and improve ML models
locally.
•Customer-managed
infrastructure.
•Fully integrated with on-
premises Orchestrator.
•Robots can be connected
to the on-prem Orch.
•Allows customers to
benefit from ALL of the
advantages of AI Center in
the cloud, without having
to migrate their existing
on-premises Orchestrator.
•Frequent patches with
new features and bug
fixes, under the cloud
platform release cadence.
•Internet access is
required.

15. 15
Identifying AI Center
Use Cases

16. 16
AI Enables Automation ofProcesses That Include
Uncertainty
You cannot determine an
outcome with 100% certainty.
High Variability
There is too much variability for
rules based.
Unstructured Data
Information like articles,
documents, images, videos and
emails.
Property
Valuation
Loan Defaults Inventory
Forecasts
Resume
Matching
Purchase
Decisions
Language
Translation
Invoice
Extraction
Email Routing Speech to
Text

17. 17
Pick a pre-built model best suited to your business needs.
No data science background required!

18. 18
Specialized AI out-of-the-box models
to accelerate time to value
Human resources, know your
customer (KYC), employee
onboarding, tax
Supply chain, shipping,
manufacturing, retail
Finance, accounts
payable, accounts
receivable, procurement
Lending, banking
Insurance, healthcare, claims
processing
Passports Bill of Lading Invoices Generic Bank Statements
ACORD 25 Certificate of Liability
Insurance
Utility Bills
Certificate of Incorporation/Good
Standing
Invoices Japan
FM 1003 Uniform Residential Loan
Application Form
ACORD 125 Commercial Insurance
Application
W2 Wage and Tax Statement Certificate of Origin Invoices China Financial Statements
ACORD 126 Commercial General
Liability Insurance Section Form
990 Return of Organization Exempt from
Income Tax-Preview
Packing Lists Invoices India Checks
ACORD 131 Umbrella / Excess
Insurance Section Form
W9 Request for Taxpayer Identification
Number and Certification
Children Product Certificate Invoices Australia Vehicle Titles
ACORD 140 Property Insurance
Section Form
1040 US Income Tax Form EU Declaration of Conformity Purchase Orders
CMS 1500 Health Insurance Claim
Form
ID Cards and Driver’s Licenses Shipping Invoices Receipts UB04
4506-T Request for Transcript of Tax Return Remittance Advice
I9 Employment Eligibility Verification Form
Generic Document Understanding model to build a model from scratch for any document type, and more Specialized AI models:
The list is not exhaustive. All the models can be trained on your own data to improve their performance for your specific requirements.
Communications Mining large language model (LLM) to extract data and sentiment from emails and other types of messages

19. 19

20. Closing message goes here.