Dell Moving AI solutions from concept into production User Manual
Moving AI solutions from concept
into production
Follow a proven path to the development
and deployment of articial intelligence
applications that solve business problems.
ABSTRACT
In this white paper, we walk through ve steps on the path to a successful articial
intelligence project, from developing a clearly dened use case to scaling a solution
in a production environment. In addition, we explore a case study highlighting the
experiences of a global enterprise that worked with the Dell Technologies HPC & AI
Innovation Lab to develop a proof of concept for an AI solution that stops credit card
fraud in real time.
Armando Acosta, Product Management
September 2020
DELL TECHNOLOGIES WHITE PAPER
TABLE OF CONTENTS
A STRATEGIC APPROACH TO AI PROJECTS . . . . . . . . . . . . 1
FIVE STEPS TO A SUCCESSFUL AI PROJECT. . . . . . . . . . . . 2
Step 1. Clearly dene the use case. . . . . . . . . . . . . . . . . . . 2
Step 2. Clean, tag, transform and catalog data . . . . . . . . . . . . . 2
Step 3. Understand your data science skillset and needs
plus train your model . . . . . . . . . . . . . . . . . . . . . . . . . 3
Step 4. Use DevOps for the last mile . . . . . . . . . . . . . . . . . . 3
Step 5. Scale to production . . . . . . . . . . . . . . . . . . . . . . . 3
A CUSTOMER CASE STUDY . . . . . . . . . . . . . . . . . . . . . . 4
Use case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Data science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
DevOps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
KEY TAKEAWAYS. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
TO LEARN MORE . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
The information in this publication is provided “as is.” Dell Inc. makes no representations or warranties of any kind with respect to the
information in this publication, and specically disclaims implied warranties of merchantability or tness for a particular purpose.
Use, copying and distribution of any software described in this publication require an applicable software license.
Copyright © 2020 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, Dell Technologies, EMC and other trademarks are
trademarks of Dell Inc. or its subsidiaries. Intel, Intel Logo are trademarks of Intel Corporation in the U.S. and/or other countries.
Other trademarks may be the property of their respective owners.
Dell Technologies believes the information in this document is accurate as of its publication date. The information is subject to
change without notice.
Published in the USA 9/20.
DELL TECHNOLOGIES WHITE PAPER
A STRATEGIC APPROACH TO AI PROJECTS
The process of developing and deploying articial intelligence applications brings unique
challenges. Complexity runs through the entire project, from dening use cases and
preparing data to training algorithms, developing software code for last mile execution and
scaling the solution in a production environment.
These are all familiar challenges to the data scientists, computer scientists and subject
matter experts at Dell Technologies. The HPC & AI Innovation Lab team regularly works
with enterprises, universities and public entities on the development of articial intelligence
algorithms and the high performance computing systems that run them.
As part of its mission, the team works proactively to share its learnings and best practices
with the global user community. One of these overarching lessons is that organizations on
the path to AI need to think strategically, across the entire lifecycle of the project, and then
move down the development and deployment path in a stepwise manner.
With that thought in mind, we suggest these ve steps toward a successful AI project:
1. Clearly dene the use case
2. Clean, tag, transform and catalog data
3. Understand your data science skillset and needs plus train your model
4. Use DevOps for the last mile
5. Scale to production
In this white paper, we walk through each of these steps, and then explore the case study
an AI project launched by a global enterprise. This organization worked with the HPC
& AI Innovation Lab to develop a proof of concept that proved the business value of an
ambitious eort to use AI to ght credit card fraud in real time.
1 DELL TECHNOLOGIES WHITE PAPER
Figure 1. rDE deployed as data proxy for Cassandra databases – conceptual architecture
Loading...