Explore how to best leverage the latest of AI technology in our upcoming webinar, where we delve into advancements and trends in the field since our previous AI webinars in 2023. Join us for a session filled with fresh insights and practical knowledge. We're stitching together the final threads of this presentation as we speak, keeping pace with AI's breakneck speed. Expect a session brimming with the freshest insights, releases and breakthroughs in AI – right up to the minute! A spotlight of this session is set to include Dmitri Bagh’s exploration of innovative AI integrations with FME, ranging from generating 3D features for augmented reality using Dall-E, to enhancing urban planning with orthoimagery completion, and showcasing the power of AI in workspace analysis and geoart creation.
Whether you're new to AI or an experienced practitioner, this webinar is tailored to keep you at the forefront of AI innovation. Get ready for a session that is as informative as it is inspiring, equipping you with the tools to excel in the dynamic world of artificial intelligence.
8. Learn how to harness the latest
AI advancements to drive
innovation and efficiency in
your projects.
Introduction
9. One platform, two technologies
FME Form FME Flow
Build and run data workflows Automate data workflows
FME Flow Hosted
Safe Software managed instance
fme.safe.com/platform
FME Enterprise Integration Platform
Safe & FME
11. Number
of
supported
data
types
in
FME
1995 2000 2005 2010 2015 2020 2023…
10
100
300
500
GIS
CAD
Database
XML
Raster
3D
BIM
Web
Point
Cloud
Cloud
Big
Data
IOT
Gaming
BI
Indoor
Mapping
AR/VR
Generative
AI
Cloud
Native
Tabular
Unrivalled Data Support
Introduction
12. 1 2 3
AuthorAI
Leveraging AI to build
workflows in FME Form
AdminAI
Leveraging AI + FME Flow
Automations to monitor
system performance
User
Consumers of FME output(s),
supplemented by AI (i.e.
reporting, analysis, etc.)
3 Themes
Introduction
13. AI is moving so quickly.
Common challenges faced:
● Keeping up with the rapid pace of AI advancements
● Applying the latest AI technologies practically
● Navigating the vast landscape of AI innovations
● Leveraging AI for real-world applications
Introduction
14. FME serves as a bridge to the world of AI
ensuring you’re always at the forefront of
advancements, simplifying the complexity of
integrating & applying latest AI developments
into your projects.
Introduction
18. Goal Block Key Result
Detect potential
bottlenecks and
automate log file
scouring
Monitoring System Performance with
Automations + AI
Log files, though
detailed, are
occasionally
difficult to read
API Driven
workflows using
FME Flow
Automation
Regular job log
auditing &
performance
monitoring
FME Flow AdminAI
22. Slide Title
Integrate with
AWS Bedrock
Foundational
Models (FMs)
Goal Block Key
Accessing high-performing foundation models
(FMs) with FME
Result
FME does not
have a native
AWS Bedrock
integration
Leverage the
bedrock-runtime
Python API in an
FME Workspace
Streamlined
integration with
industry-leading
Foundational
Model providers
24. Virtually limitless
integration possibilities
Pre-built connectors allow you to easily connect to AWS and
Azure based services. FME can also easily be extended via
REST, SOAP, and Python based API’s to connect to even more
services & models.
25. We strive to make it easier to
connect to more [AI &
Business] systems than ever
before
26. FME enables easy/re-usable integration with
native transformers and various API
integration capabilities:
● REST/SOAP: HTTPCaller
● Python: PythonCaller/PythonCreator
Code-Optional Workflows
Amazon Bedrock & FME
27. Each FM has a specific body/response
structure.
● FME Transformers enable users to easily
format upload bodies & extract response
bodies
○ Replace Strings, Flatten/Fragment
JSON/XML bodies, etc.
Resource: Getting Started with JSON, Getting Started with XML
Code-Optional Workflows
Amazon Bedrock & FME
28. Enable more FME users to leverage your AI
connector
● Create your own Custom Transformer
● Share it on the FME Hub
● Leverage Published Parameters to make
it easily re-usable
○ Parameter Manager
Creating Custom Transformers
Amazon Bedrock & FME
31. Integrate any AI Service Provider
with FME
● Automate troubleshooting & monitor system performance
● Generate summary content for reporting
● Automate transcription & action items
FME & AI Services
34. ● Explore New Use Cases
○ Project scope & time estimation
● Troubleshooting & Debugging
○ Search and suggest resolution for
error message(s)
● Learning & Skill Development
○ Explain functionality & how to build a
workspace*
* Always use your best judgement as AI can still produce inaccurate
results. The FME Community & Documentation are great for fact checking.
Opportunities for AI outside of
FME
Co-Authoring Workspaces
35. Enter your project requirements
Exploring New Use Cases
Co-Authoring Workspaces
36. Enter your project requirements
● AI lists requirements
Exploring New Use Cases
Co-Authoring Workspaces
37. Enter your project requirements
● AI lists requirements
● Suggests key steps to building the
Automation
Exploring New Use Cases
Co-Authoring Workspaces
38. Enter your project requirements
● AI lists requirements
● Suggests key steps to building the
Automation
● Provides Time Estimate and training
resources
Exploring New Use Cases
Co-Authoring Workspaces
39. Pull errors directly from the log file & paste
them in the chat
● Suggests alternative code snipped
● Provides insight into a best practice
● Adds suggestion for handling HTTP
related errors
Troubleshooting & Debugging
Co-Authoring Workspaces
40. Explain ETL concepts & supplement your
Code Optional components with AI feedback
● Enhance your HTML reporting output for
Data Streaming Jobs
● Describe a task to get recommended
transformers to use
* Sample HTML features were randomly generated by OpenAI and
not based on Census data
Learning & Skill Development
Co-Authoring Workspaces
41. 5
OpenAI API as a part
of FME workflows
Now we can enhance the transformation outcomes by
combining the power of FME and artificial intelligence
42. OpenAI API as a part of FME workflows
Previously demoed
scenarios
● Texturing with Dall-E
● Projection selection
● 3D geometry coloring
● FME Log messages to Slack voice messages
43. OpenAI API as a part of FME workflows
New scenarios
● Text to 3D transformations
● Workspace Analysis
● Filling Gaps on Orthoimagery for Urban Planning
● Batch Files for File Management
● GeoArtAI
44. Text to 3D transformations
● Specialized AI models are in their infancy
● Currently, very narrow applicability
● But it worked!
Read more here
OpenAI API as a part of FME workflows
45. Text to 3D transformations
OpenAI API as a part of FME workflows
46. Workspace analysis
● Feed a screenshot of a workspace to
ChatGPT (or this service) and ask what
the workspace does
● Ask your colleague the same question
😉
● Compare who does it better
Read more here
OpenAI API as a part of FME workflows
49. Filling Gaps on Orthoimagery for
Urban Planning
● Use FME to clip imagery by parcel, block,
zoning, etc.
● Use Dall-E2 model to fill the gaps
● Generate up to 10 variations with a single
creative prompt
Read more here
OpenAI API as a part of FME workflows
50. Filling Gaps on Orthoimagery for
Urban Planning
● Use FME to clip imagery by parcel, block,
zoning, etc.
● Use Dall-E2 model to fill the gaps
● Generate up to 10 variations with a single
creative prompt
OpenAI API as a part of FME workflows
51. Scripting for File Management
● Type what you want to do with your files
● Run the workspace
● Done!
Read more here
OpenAI API as a part of FME workflows
52. GeoArtAI
● Building parametrized prompts for
creating prompts 😀
● HTML/CSS/JS/Leaflet with ChatGPT
● FME with JSON writer for data delivery
● Geospatial Bias is real!
● GeoArtAI and Gallery
Read more here
OpenAI API as a part of FME workflows
54. OpenAI API as a part of FME workflows
Let ChatGPT create HTML/CSS/JavaScript
55. OpenAI API as a part of FME workflows
Two workspaces powering the service
56. OpenAI API as a part of FME workflows
Geospatial Bias
Geospatial Bias refers to a computational bias where there is a predisposition to assume or
prioritize certain geographic locations over others in the absence of explicit information, or
misinterpretation or incorrect prioritization of specific locations despite clear and precise
geographic data. (Read more here)
57. OpenAI API as a part of FME workflows
[Blog] AI and FME: 12 Inspiring Scenarios for Data Integration
59. Get our Ebook
Spatial Data for the
Enterprise
fme.ly/gzc
Guided learning
experiences at your
fingertips
academy.safe.com
FME Academy
Resources
Check out how-to’s &
demos in the knowledge
base
community.safe.com
/s/knowledge-base
Knowledge Base Webinars
Upcoming &
on-demand webinars
safe.com/webinars
61. We’d love to help you get
started.
Get in touch with us at
info@safe.com
Experience the
FME Accelerator
Contact Us
A world where data is not just a
commodity but a catalyst for
real change.
fme.safe.com/accelerator/
Next Steps
62. ClaimYour Community Badge
● Get community badges for watching
webinars!
● fme.ly/WebinarBadge
● Today’s code: LCBGA
*** Please hold onto this code, redeem after
new FME Community launch mid-month!
Join the Community today!
Next Steps