UiPath Community Day Amsterdam presentations

Welcome to our third live UiPath Community Day Amsterdam! Come join us for a half-day of networking and UiPath Platform deep-dives, for devs and non-devs alike, in the middle of summer ☀.

📕 Agenda:
12:30 Welcome Coffee/Light Lunch ☕
13:00 Event opening speech
Ebert Knol, Managing Partner, Tacstone Technology
Jonathan Smith, UiPath MVP, RPA Lead, Ciphix
Cristina Vidu, Senior Marketing Manager, UiPath Community EMEA
Dion Mes, Principal Sales Engineer, UiPath

13:15 ASML: RPA as Tactical Automation
Tactical robotic process automation for solving short-term challenges, while establishing standard and re-usable interfaces that fit IT's long-term goals and objectives.
Yannic Suurmeijer, System Architect, ASML
13:30 PostNL: an insight into RPA at PostNL
Showcasing the solutions our automations have provided, the challenges we’ve faced, and the best practices we’ve developed to support our logistics operations.
Leonard Renne, RPA Developer, PostNL
13:45 Break (30')
14:15 Breakout Sessions: Round 1
Modern Document Understanding in the cloud platform: AI-driven UiPath Document Understanding
Mike Bos, Senior Automation Developer, Tacstone Technology
Process Orchestration: scale up and have your Robots work in harmony
Jon Smith, UiPath MVP, RPA Lead, Ciphix
UiPath Integration Service: connect applications, leverage prebuilt connectors, and set up customer connectors
Johans Brink, CTO, MvR digital workforce
15:00 Breakout Sessions: Round 2
Automation, and GenAI: practical use cases for value generation
Thomas Janssen, UiPath MVP, Senior Automation Developer, Automation Heroes
Human in the Loop/Action Center
Dion Mes, Principal Sales Engineer @UiPath
Improving development with coded workflows
Idris Janszen, Technical Consultant, Ilionx
15:45 End remarks
16:00 Community fun games, sharing knowledge, drinks, and bites 🍻

3. Leonard Renne
18
th
of July 2024
An insight into RPA at
PostNL

4. 4
•Introduction
•RPA at PostNL
•HR: SuccessFactors
•Machine Learning
•Best Practice
•Challenges we face
•What’s next?
Agenda

5. Introduction
5
Leonard Renne

6. RPA at PostNL
6
•Proof of concept
•Building infrastructure
•First bots
Current situation
•Two RPA teams
•100+ Bots in production
•Almost 200.000 hours
Start

7. 2024
•APIs
HR –SuccessFactors
7
2019-2020
•2 Developers
•4 Releases p/y
2022
•Descriptors
•2 Releases p/y
2021
•Launch of preview
environment
2023
•Libraries

8. Machine Learning
8

9. Best Practice
9
Testcases

10. Challenges we face
10
•Finding the right projects
•Priority
Who can relate?

11. What’s next?
11
•Fluctuations of usage
•Testing
Generative AI
•ChatGPT
•Alternatives
Robot as a Service

12. Thank you for
your time!

13. 24 July 2024 13

14. 24 July 2024
Fastertraining, no-code
PoweredbySpecializedandGenerativeAI
Modern document understanding

15. 24 July 2024
•Work at Tacstone Technology since January 1, 2020
•Co-Guild Lead of uipathdevelopmentteam
•Established our working methodology with Azure DevOps
•FormerUiPath MVP (2023)
•Certified UiARDand UISAI, currently working on the 'ASAP’ Solution Architect certification
WhoamI?
15 https://www.linkedin.com/in/mike-bos
Mike Bos
Lead Developer at Tacstone Technology

16. 24 July 2024
DU & You
16
Whohas experiencewithDocument Understanding?
•mentimetet

17. 24 July 2024
Whatis Modern DU?
17
Active Learning*
•Next-gen AI-poweredmodel training experience (AI + Gen AI)
•Real time Iterative processbetween annotators and the model
•Automatic annotationsuggestions
No-code
Guidanceon model optimization
Automatic classification
*Modern DU + AL: Retrainingfromuser feedback availablesoon!

18. 24 July 2024 18
Generativevalidation
UsesGen. AI tovalidateextractionresultswitha confidencebelow threshold
Ensureshigheraccuracy
Increasingefficiency
Reducesneedforhuman interaction
Simplyrequiresactivatingthecheckmark in the‘Data ExtractionScope’ or in ‘Extract Document Data

19. 24 July 2024
No pre-trainedmodel needed
No confidencescore, impactinghitllogic.
Easy extractors:
•"The total amount of all the items on the receipt as decimal number. Only extract
numbers and decimal points. make sure to not select the sub total.“
Demo
GenerativeExtractor
19

20. 24 July 2024
Compared to IntelligentOCR, there are some missing features, which are
planned to be added. Some examples are:
Splitting documents
Business Rules
Training (fine-tuning models)
Tobeadded
20

21. 24 July 2024
AI-unit consumption
Product Unit cost
Infrastructure
Model training
Consumption
No additionalcosts
No additionalcosts
Anycombinationof OCR,
Classificationandextraction
costs1 AI Unit per pagein
total
Optional: Using generative
validationcostsanadditional
AI Unit per page
AI-units availableper 60k units
21

22. 24 July 2024
Createmodern project
•Build
•Upload & Classify
oCustom
oReceipts
oInvoices
•Document type manager
oFields
oSettings
oAnnotate
•Measure
•Project Score
•Factors
•Metrics
Demo
22
•Publish
•Version
•Studio web
•API
•Monitor
•Time saved
•Cost
•Processeddocuments
•Apply(Studio)
•Adddocstobucket
•Extract Document Data
•Proces
•Extract Document Data (Generative)
•DocumentDataContents

24. Process Orchestration
Scale up and have your Robots work in harmony

25. 26
Introduction
Ciphix
Jon Smith
RPA Team Lead / Solution Architect

26. 01
02
03
Chapters

27. Use Case / Example

28. 29
Use Case
GVRT Reporting
Upload Correctsions
File
Generate Returns
Run VAT Checks
Run Tests On
Transactions
Explain Any
Transactions That
Don’t Pass
Upload To Financial
Reporting System

29. 30
Sequential Design
If We Design Using Performer Chaining
Performer 1
Check VATs
Run Tests On Transactions
Performer 2
Upload Corrections
Generate Returns
Human In The Loop
Somehow
Performer 3
Upload To Reporting
System

30. 31
Sequential Design
Difficulty When Inserting A Change
Make Queue Item
For Performer 2
Make Queue Item
For Performer 3
Performer 2 Performer 3Performer 1

31. 32
Sequential Design
Difficulty When Inserting A Change

32. Orchestration Concept

33. 34
Terminology
What are you talking about?

34. 35
Terminology
Two Main Types
Business Orchestration
High Level
Multiple Distinct Processes
Integration Of Business
Units
Optimizations Across The
Organization
Process Orchestration
Focused
Many SmallerTasks
Combined
Multiple Applications*
Human In The Loop*

35. 36
Why Change From Sequential Processing?
One may break the chain, or can we keep going?

36. 37
Why Change From Sequential Processing?
One may break the chain, or can we keep going?
•Less time to create*
(*on first sight)
•Easy to report on*
(*on first sight)
•Less parts
(*on first sight)
•Better resource management
•Better information control
•Easier scalable
•Better error mitigation
•More flexible
Sequential Processing Orchestration

37. 38
Why Change From Sequential Processing?
Orchestration Process Facilitates Easier Changes
Orchestration Process
Performer 1 Performer 2 Performer 3

38. How To Do It In UiPath
With Persistence

39. 40
Process Orchestration
The current setup
Performer 1
Check VATs
Run Tests On Transactions
Performer 2
Upload Corrections
Generate Returns
Human In The Loop
Somehow
Performer 3
Upload To Reporting
System

40. 41
Process Orchestration
Its all about Persistence
Trigger
Performer 1
Human In The
Loop
Trigger
Performer 2
Performer 1
Check VATs
Run Tests On Transactions
Performer 3
Generate Returns
Orchestration Process
Yes No
Performer 2
Upload Corrections
Trigger
Performer 3
Trigger
Performer 4
Performer 4
Upload Results
Transactions Require
Explanation?

41. 42
Process Orchestration
Seeing It In Studio

42. 43
Process Orchestration
Flexible To Changes
Trigger
Performer 1
Human In The
Loop
Trigger
Performer 2
Performer 1
Check VATs
Run Tests On Transactions
Performer 3
Generate Returns
Orchestration Process
Yes No
Performer 2
Upload Corrections
Trigger
Performer 3
Trigger
Performer 4
Performer 4
Upload Results
Transactions Require
Explanation?

44. Integration Services
UiPathDEV meet-up
Jeroen Backx
Johans Brink
18-7-2024

45. Johans Brink
CTO
[email protected]
Jeroen Backx
Senior Consultant
[email protected]

46. Component Relationships

47. Add connection cloud / workspace

48. Oauth/ Other app

49. Admin consent Azure

50. Connection Builder

52. Authentication

58. Questions

60. Automation and GenAI:
practical use cases for
value generation
18 July2024

61. 62
•Thomas Janssen
•Founderof Tom’s Tech Academy
•Full time RPA trainer
•+600K views on my YouTube channel
every year
WhoamI?

62. 63
Extract information fromemail witha cloudLLM
Functioncalling
UseRAG toansweremails
Today’sagenda:
I hope to inspire you with this new technology ☺
This is innovation, and far from production ready

63. 64
GenAIUseCases: Classification
Generatecontent
Generateinsights
Interact

64. Extract information fromemail witha cloudLLM
Usecase 1

65. 66
Usecase 1: Email Info Extraction

66. 67
Usecase 1: Email Info Extraction

67. 68
Use case 1: Email Content Extraction
"You receive emails from customers. You'll respond in valid JSON in the format I
provide. You respond in JSON ONLY.
For every email you'll provide the following response:
{'InvoiceQuestion':True,'SuccesfulExtraction':True,'CustomerNumber':'12234','InvoiceNu
mber':'123'
In the first field InvoiceQuestionyou classify whether this question is about an invoice.
You answer either True or False.
In the second field SuccesfulExtractionyou tell me whether you could succesfullyextract
BOTH the invoice number and the customer number (True or False)
In the third field you put the Customer Number
In the fourth field you put the Invoice Number
If either one of the first two questions is answered with False, you don't provide the
invoice and customer number
}
"

68. 69

69. 70

70. 71

71. Agent withFunctionCallingcapabilities
Usecase 2

72. 73
Fetch
occasions from
database
Search
occasion for…
Proposea
price..
Booktest ride
Buythecar

73. 74
Usecase 2: Functioncalling

74. Retrieval AugmentedGeneration
Usecase 3

75. 76
GenAI’sproblem
No specific data about your
company
Information is outdated or simply
wrong

76. 77
The solution
User Librarian
Book

77. 78
Retrieval AugmentedGeneration(RAG)

78. 79
Usecase 3: Retrieval AugmentedGeneration

79. 80
Thankyou!
Interested in working together?
https://tomstechacademy.nl/

80. 81

81. UiPath Action Center
Engagehumans-in-the-loop

82. 83

83. How does Action Center
help Automation

84. Action Center for business users
Achieve business continuity
Robots available for jobs,while humans take actions​
Improved decision making​
Easy exception and escalationmanagement
Bulk handling of actions
Action summary dashboard
Easily categorize actions
Real-time notifications, in-ap-and email
Metadata on users and actions
Custom, global notification experience

85. Action Center for RPA developers
Long-running workflows that work with cross-platform robots
Wider range of processes can be automated
Simplified compliance, centralized coordination
Attain end-to-end visibility ofbusiness processes
Configure labels, label data and export labeled data for ML models

86. Use cases

87. 88
Approval Verification Validation
Humans review documents against a checklist,
identify the missing documents, and attach the
necessary documents
Humans approve or reject something, based
on which a workflow take a particular path
Humans validate ML model predictions. Robots
send human-validated data back forretraining
Use case examples
New vendor onboarding
Customer onboarding
Account activation
Loan forgiveness
T&E audit compliance
Invoice processing
GL coding in invoices
Order management
Vendor management
Receipt processing
Invoice processing
Purchase order processing
Email classification
Image classificationPayment operations
•For more use case examples, please check out the use casesection of this deck.

88. 89
Approval scenarios
A business user is presented with a snapshot of a business context from a workflow to approve or reject
something, based on which the workflow takes a particular path for completion.
Apply country rules
to devise discount
Request for quotesFetch line-item details
Approver : 1
Auto approved
Quote generation
and dispatch
Discount > 10%
Approver : 2
Human Attended Robot Unattended Robot

89. Verification scenarios
Document verification process awaiting on agents to review something against a checklist, a pre-set business rule, or
based on knowledge. Humans verify, identify missing documents (business exceptions), attach necessary documents,
make decisions, or escalate.
•Examples: new vendor onboarding, customer onboarding, account activation, invoice reconciliation, loan application review
Robot finds missing
documents, and inform
agent
Loan package read
from the queue
Check document
completeness
against a check list
Agent verifies, uploads
missing documents
and marks as complete
All documents in place,but
loanamountisbiggerthana
certain amount
Robot completes
and sends email
notification
Human Attended Robot Unattended Robot
Agent verifies and
marks review as
complete
All documents are in place,
andtheloanamountis
lower than a certain amount

90. Validation scenarios
Business user performs validation of machine-extracted data based on a threshold. Action Center can be used to easily
build custom UI for business users to validate ML model predictions.
UiPath
SAP
Monitor emails
and download
attachments
Extract data
from
documents Pre-validation
SAP MDM
User validates the
order content
Date and
quantity checks
Sales order
preparation
SAP sales
order
processing
SAP order
confirmation email to
the customer
Sales order Update
HITL with Task inbox
Human Attended Robot Unattended Robot
•Customer stories: Cognizant, Evros, Accelirate, Amitech, Pepsi, Electrocomponents

91. 92
Actions: Simplify collaboration with
robots through the Actions inbox
Centralized inbox accessible through web-based
portal (outside of Orchestrator) and mobile app
Exception, validation, escalation,
andapprovalhandling byusers
User access management,dynamic
taskassignmentand grouping
Automated task creation by robots and resumption
ofworkflows after human action is performed​

92. 93
Action Center
Global Notification Service:Near to real-time, email & in-app notificationsfor ActionCenterevents​

93. 94
Action Center
Actions summary dashboards: Business admins & users have a consolidated view of pending
and completed Actions in an overview page and improved categorized views for planning the
work

94. 95
Action Center
Actions handling within Assistant: Users can now access and handle Actions from a native
digital assistant

95. Integration

96. 97
Advanced monitoring dashboards Actionable insights
Receive a notification in Action Center whenever
exception handling is required in the process, or
trigger Robots to handle idle tasks.
User-oriented dashboards, KPIs and tags to closely
monitor automation and process outcomes
Action Center + Process Mining

97. 98
Action Center + Document Understanding
Load
taxonomy
to define
document types
and fields for
processing
Digitize
documents using
OCR to make
them machine-
readable
Export
the extracted
data for further
usage
Classify
and split the files
into document
types
Validate
classification
results (human
review)
Train
classifiers
based on the
validated data
Extract
information from
the documents
Validate
extraction results
(human review)
Train
extractors
based on the
validated data

98. Demo

99. Two processes working together
Bot initiates task
to select
outstanding
invoices
User assigns
task to
themselves
User selects
invoices in Task
App to send
reminders
App starts
processes to
send reminders
Bot creates
email text
based upon
invoice data
Bot creates
task to
approve and
adjust email
User assigns
task to
themselves
User accepts
text
Bot sends
email

100. Activities

101. 102
•Start Job And Get Reference
•Wait For Job And Resume
•Add Queue Item And Get Reference
•Wait For Queue Item And Resume
•Create Form Task
•Wait For Form Task And Resume
•Resume After Delay
•Assign Tasks
•Create External Task
•Wait For External Task And Resume
•Complete Task
•Forward Task
•Get Form Tasks
•Get Task Data
•Add Task Comment
•Update Task Labels
•Create App Task
•Wait For App Task And Resume
•Configure task timer
Activities