The Operations effect of Dev, Data, ML and FM

In today's rapidly evolving tech landscape, the convergence of various "Ops" practices is revolutionizing how organizations develop, manage, and scale their applications, data, machine learning, and facilities. This session will delve into the origins of four critical "Ops" disciplines—DevOps, DataOps, MLOps, and FMOps—and explore the exciting paths they're charting for the future.

1. PUNE 2023

2. The Ops effect of Dev,
Data, ML and FM
Bhuvaneswari Subramani | Oct 2023
PUNE 2023

3. #WhoAmI
Bhuvaneswari Subramani
@installjournal
https://www.linkedin.com/in/bhuvanas/
https://www.bhuvana.pro/
Chief Cloud Evangelist @ Intuitive.Cloud
AWS Hero
Technologist, Global speaker and passionate blogger

4. PUNE 2023
DevOps - Pioneering Agility
DataOps - Streamlining Data Pipelines
01
02
03MLOps - Enabling AI at Scale
04FMOps - Empowering the Future of Gen AI Apps
Key Insights

5. What's your all-time favourite ice cream flavour?
The Scoop Challenge
Do you prefer classic ice cream cones or innovative ice cream
creations?
Are you adventurous when it comes to trying new and exotic ice
cream flavours?
Do you have any nostalgic ice cream memories from your
childhood?

6. PUNE 2023
The Sweet Journey of
ChillDelights Creamery

7. Slow New
Flavour
Development
Disconnected
Customer
Feedback
Unpredictable
Favour
Success
Manual
Customer
Support
Challenges
ChillDelights Creamery

8. Before DevOpsAfter DevOps
Image taken from freepik.com

9. Before DataOpsAfter DataOps
Image taken from freepik.com

10. Before MLOpsAfter MLOps
Image taken from freepik.com

11. Before FMOpsAfter FMOps
Image taken from freepik.com

12. DevOps – Pioneering Agility
PUNE 2023

13. DevOps in ChillDelights Creamery
Dev
Ops
•Rapid Feature Deployment
•Efficient Testing and Validation
•Collaborative Franchise Support

14. Benefits of DevOps
Speed
Rapid Delivery
Reliability
Scale
Security
Improved Collaboration

15. DevOps LifeCycle
CODE
PLAN
BUILD
RELEASE
DEPLOY
OPERATE
Dev
Ops
TEST
MONITOR

16. AWS Services in DevOps Lifecycle
CODE
PLAN
BUILD
RELEASE
DEPLOY
OPERATE
Dev
Ops
TEST
MONITOR
AWS CodeDeploy
AWS CodeBuild
AWS
CodeCommit
Cloud 9
Amazon
CloudWatch
Third Party
Tooling
AWS Elastic
Beanstalk
AWS
CodePipeline
AWS
CloudFormation

17. What happens when there is no CI/CD ??

18. DevOps Pipeline for Continuous Integration

19. DevOps Pipeline for Continuous Deployment

20. Harness Pro
DevOps Tools on AWS Marketplace
Dynatrace

21. DataOps - Streamlining Data
Pipelines
PUNE 2023

22. Mystery without DataOps

23. 5 V’S OF BIG DATA
More data,
better insights,
accurate
predictions
VOLUMEVELOCITY
VARIETY
VALUEVERACITY
Data..
Real-time
decisions, swift
responses,
competitive
advantage
Diverse data,
richer insights,
comprehensive
analyticsHigh quality,
trust, reliable
decision-making
High revenue,
cost savings,
innovation,
improved
performance

24. DataOps Lifecycle

25. Benefits of DataOps
Collaboration
Increased efficiencies
Reduced costs
Orchestration
Data Security
Data Reusability

26. DataOps services in AWS
Amazon Kinesis
Amazon EMRAmazon AthenaAmazon RedshiftAWS Glue
AWS
Data Pipeline
Amazon QuickSight

27. DataOps Pipeline for Big Data Analysis

28. MLOps - Enabling AI at
Scale
PUNE 2023

29. Are you on cross-roads?

30. Lifecycle of MLOps

31. Benefits of MLOps
Scalability
Model Versioning
& Tracking
Security &
Compliance
Automation
Collaboration &
Integration
Cost Efficiency

32. MLOps Services in AWS
Amazon
SageMaker
Amazon
Comprehend
Amazon
Personalize
Amazon
Rekognition
Amazon
Forecast
Amazon
Transcribe

33. MLOps Pipeline for Personalized Recommendations

34. FMOps - Empowering the
Future of Gen AI Apps
PUNE 2023

35. Put yourself into the shoes of the Customer Service Team?
It’s taking too long to reply to
all customers’ messages
Please
HELP!! CHATBOT
at your
service
24*7

36. MLOps vs FMOps
Traditional MLOps
•Combines multiple
task-specific models
and business logic
downstream.
•Not well-suited for FMs
due to their complexity
and opaqueness.
FMOps in the age of FMs
•Focuses on upstream
smart data preparation,
fine-tuning and
guidance of emergent
FM behavior, and
further postprocessing
and chaining of FM
outputs.
•Necessary because FMs
are trained on massive
datasets and are able
to learn complex
patterns in data.

37. Lifecycle of FMOps

38. Benefits of FMOps
Reliability
Security
Reduced costsImproved
Performance
Better
Collaboration

39. FMOps Services in AWS
AWS GlueAmazon SageMakerAmazon Translate
Amazon ComprehendAmazon Bedrock

40. FMOps Pipeline for ChatBot

41. The Scoop
DevOps – Agility into POS Software features
DataOps - Derive Insights from Data
MLOps – Personalization, Forecasting & Innovation
FMOps - Your personal assistant at reach

42. The Ops Effect continues..

43. Chief Cloud Evangelist @ Intuitive.Cloud
Bhuvaneswari Subramani
Advance your career & achieve your
professional goals!!
Follow me to receive insights on
cloud modernization, FinOps, DevOps, & leadership
https://intuitive.cloud/careers
Work on the next
BIG THING at Intuitive