SlideShare a Scribd company logo
1 of 32
IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation 1
How to deploy machine
learning models into
production
Sumit Goyal (@SumitG0yal)
Software Engineer, Watson Studio
Sven Hafeneger
Data Scientist, Watson Studio
IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation 2
About me
Sumit Goyal (@SumitG0yal )
Software Engineer, Watson Studio
Agenda
3IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Motivation
4IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining
Ops
Dev
Motivation
5IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining
Work does not end with
handover of ML models !
Motivation
6IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation http://pobsonline.com/money-mistakes/, https://www.drwf.org.uk/news-and-events/events/100-day-challenge
Mike
Data Scientist
Deb
The Developer
ModelDev Ops
Production Environment
Motivation
7IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
,http://pobsonline.com/money-mistakes/, https://www.drwf.org.uk/news-and-events/events/100-day-challenge
, https://de.wikipedia.org/wiki/Zahnrad
Mike
Data Scientist
Deb
The Developer
Model + App
DevOps
Production Environment
Motivation
8IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining
• Production environment is in the Cloud
• Python as primary programming language
• Micro-service architecture
• Python ML Frameworks
• Traditional ML algorithms, no deep learning
• Data Science Team <-> Software Development Team
Agenda
9IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
ML Model artifacts
10IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
PMML
Serialized Object
Re-Implementation
in Java/C++
Python ML Model Production Environment
App
REST
API
App
ML Model artifacts
11IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
PMML
Serialized Object
Re-Implementation
in Java/C++
App
REST
API
App
Python ML Model Production Environment
Handover model, code, integration
test, no re-train in Travis
Agenda
12IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Production Environments
13IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Serialized Object
(Python)
Meta data
- Version
- Name
- Type
- ML Framework
- Eval. Metrics
- …
Python Application - REST API
Endpoint: /v1/xyz/prediction : POST
Webserver / Web
framework
Endpoint
ML framework
- Load ML model
- Create a route
- Prepare data
- Run prediction
- Post-processing
- Return result
Logic
https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html
Production Environments – eCommerce
14IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Endpoint
REST API
App
Service agreement
- max. response time
- availability
- quality/confidence
of prediction
- max. re-train time
- monitor
- …
https://blog.paymill.com/en/keep-e-commerce-system-functioning-optimum-capacity/,
http://www.atlascorps.org/blog/customers-and-clients-the-devils-of-business-growth/
App
App
Production Environments – eCommerce
15IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Endpoint
REST API
App
Service agreement
- max. response time
- availability
- quality/confidence
of prediction
- max. re-train time
- monitor
- …
https://blog.paymill.com/en/keep-e-commerce-system-functioning-optimum-capacity/,
http://www.atlascorps.org/blog/customers-and-clients-the-devils-of-business-growth/
App
App
Order == Fraud ?
Order!
Production Environments – Fraud Detection
16IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://www.kaggle.com/mlg-ulb/creditcardfraud
Ordering Transactions
Agenda
17IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Deployment with Cloud Foundry
18IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Python App - REST API
Endpoint: /v1/xyz/prediction : POST
REST
- Load ML model
- Create a route
- Prep. score data
- Run prediction
- Post-processing
- Return result
Model
https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org
Questions
- How to configure the app ?
- How many instances ?
- How much RAM ?
- Zero downtime deployment ?
- Connection to DB?
- Logging ?
- Crashes and recovery?
- Routing ?
- Workload scaling ?
- Integration with other tools?
- …
Deployment with Cloud Foundry
19IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Choose a build pack !
Deployment with Cloud Foundry
20IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Python App: /route/..
http://flask.pocoo.org/, http://scikit-learn.org
../projects/fraud_detection_api
CLI: push CLI: set-env
Configuration
R
E
S
T
A
P
I
Code
App
POST: {body}
Order == Fraud ?
Deployment with Cloud Foundry: Demo
21IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cloud_Foundry
Agenda
22IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Deployment with Kubernetes
23IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Container - REST API
Endpoint: /v1/xyz/prediction : POST
Endpoint
- Create a route
- Load ML model
- Prepare data
- Run prediction
- Post-processing
- Return result
Logic
https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org
Questions
- How to configure the app ?
- How many instances ?
- How much RAM ?
- How many CPUs/GPUs?
- Zero downtime deployment?
- Connection to DB?
- Logging ?
- Crashes and recovery?
- Routing ?
- Workload scaling ?
- Integration with other tools?
- …
- How much do you want to
customize the runtime?
Deployment with Kubernetes
24IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
Docker Container
REST
API
Logic
Docker
Images
Repository
App
(Container)
POST: {body}
https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org, https://blog.takipi.com/pivotal-
cloud-foundry-vs-kubernetes-choosing-the-right-cloud-native-application-deployment-platform/
Order == Fraud ?
25IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Kubernetes
Deployment with Kubernetes: Demo
Agenda
26IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivation
• Python ML Model artifacts
• Production Environments
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Deployment with managed services
27IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
ML Models
Deployment
R
E
S
T
A
P
I
App
POST: {body}
Order == Fraud ?
Manages Deployment Service
Deployment with managed services
28IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://docs.microsoft.com/en-us/azure/machine-learning/, https://aws.amazon.com/sagemaker/, https://console.bluemix.net/catalog/services/machine-learning
IBM Watson Machine Learning
Deployment with managed services: Demo
29IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://console.bluemix.net/catalog/services/machine-learning
Summary
30IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Motivated and informed
• Deployment with Cloud Foundry
• Deployment with Kubernetes
• Deployment with managed services
Learn more about:
31IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation
• Cloud Foundry: https://www.ibm.com/cloud/cloud-foundry
• Kubernetes: https://kubernetes.io/
• Watson Machine Learning: https://www.ibm.com/cloud/machine-learning
Thank you for your attention.
Time for QA!
Sumit Goyal
sumit.goyal@de.ibm.com
32IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation

More Related Content

What's hot

Apply MLOps at Scale
Apply MLOps at ScaleApply MLOps at Scale
Apply MLOps at ScaleDatabricks
 
“Houston, we have a model...” Introduction to MLOps
“Houston, we have a model...” Introduction to MLOps“Houston, we have a model...” Introduction to MLOps
“Houston, we have a model...” Introduction to MLOpsRui Quintino
 
MLOps Virtual Event: Automating ML at Scale
MLOps Virtual Event: Automating ML at ScaleMLOps Virtual Event: Automating ML at Scale
MLOps Virtual Event: Automating ML at ScaleDatabricks
 
Feature Store as a Data Foundation for Machine Learning
Feature Store as a Data Foundation for Machine LearningFeature Store as a Data Foundation for Machine Learning
Feature Store as a Data Foundation for Machine LearningProvectus
 
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...Building a MLOps Platform Around MLflow to Enable Model Productionalization i...
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...Databricks
 
"Managing the Complete Machine Learning Lifecycle with MLflow"
"Managing the Complete Machine Learning Lifecycle with MLflow""Managing the Complete Machine Learning Lifecycle with MLflow"
"Managing the Complete Machine Learning Lifecycle with MLflow"Databricks
 
MLOps Using MLflow
MLOps Using MLflowMLOps Using MLflow
MLOps Using MLflowDatabricks
 
Ml ops intro session
Ml ops   intro sessionMl ops   intro session
Ml ops intro sessionAvinash Patil
 
Using MLOps to Bring ML to Production/The Promise of MLOps
Using MLOps to Bring ML to Production/The Promise of MLOpsUsing MLOps to Bring ML to Production/The Promise of MLOps
Using MLOps to Bring ML to Production/The Promise of MLOpsWeaveworks
 
Slide DevSecOps Microservices
Slide DevSecOps Microservices Slide DevSecOps Microservices
Slide DevSecOps Microservices Hendri Karisma
 
Machine Learning Data Lineage with MLflow and Delta Lake
Machine Learning Data Lineage with MLflow and Delta LakeMachine Learning Data Lineage with MLflow and Delta Lake
Machine Learning Data Lineage with MLflow and Delta LakeDatabricks
 
The Paved Road at Netflix
The Paved Road at NetflixThe Paved Road at Netflix
The Paved Road at NetflixDianne Marsh
 
Managing the Machine Learning Lifecycle with MLflow
Managing the Machine Learning Lifecycle with MLflowManaging the Machine Learning Lifecycle with MLflow
Managing the Machine Learning Lifecycle with MLflowDatabricks
 
Software estimation
Software estimationSoftware estimation
Software estimationMd Shakir
 
Introduction to GitHub Copilot
Introduction to GitHub CopilotIntroduction to GitHub Copilot
Introduction to GitHub CopilotAll Things Open
 
What Is DevOps?
What Is DevOps?What Is DevOps?
What Is DevOps?Soumya De
 

What's hot (20)

Track code quality with SonarQube
Track code quality with SonarQubeTrack code quality with SonarQube
Track code quality with SonarQube
 
Apply MLOps at Scale
Apply MLOps at ScaleApply MLOps at Scale
Apply MLOps at Scale
 
“Houston, we have a model...” Introduction to MLOps
“Houston, we have a model...” Introduction to MLOps“Houston, we have a model...” Introduction to MLOps
“Houston, we have a model...” Introduction to MLOps
 
MLOps Virtual Event: Automating ML at Scale
MLOps Virtual Event: Automating ML at ScaleMLOps Virtual Event: Automating ML at Scale
MLOps Virtual Event: Automating ML at Scale
 
Feature Store as a Data Foundation for Machine Learning
Feature Store as a Data Foundation for Machine LearningFeature Store as a Data Foundation for Machine Learning
Feature Store as a Data Foundation for Machine Learning
 
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...Building a MLOps Platform Around MLflow to Enable Model Productionalization i...
Building a MLOps Platform Around MLflow to Enable Model Productionalization i...
 
"Managing the Complete Machine Learning Lifecycle with MLflow"
"Managing the Complete Machine Learning Lifecycle with MLflow""Managing the Complete Machine Learning Lifecycle with MLflow"
"Managing the Complete Machine Learning Lifecycle with MLflow"
 
MLOps Using MLflow
MLOps Using MLflowMLOps Using MLflow
MLOps Using MLflow
 
Ml ops intro session
Ml ops   intro sessionMl ops   intro session
Ml ops intro session
 
Using MLOps to Bring ML to Production/The Promise of MLOps
Using MLOps to Bring ML to Production/The Promise of MLOpsUsing MLOps to Bring ML to Production/The Promise of MLOps
Using MLOps to Bring ML to Production/The Promise of MLOps
 
Introduction to Git and GitHub
Introduction to Git and GitHubIntroduction to Git and GitHub
Introduction to Git and GitHub
 
Slide DevSecOps Microservices
Slide DevSecOps Microservices Slide DevSecOps Microservices
Slide DevSecOps Microservices
 
Machine Learning Data Lineage with MLflow and Delta Lake
Machine Learning Data Lineage with MLflow and Delta LakeMachine Learning Data Lineage with MLflow and Delta Lake
Machine Learning Data Lineage with MLflow and Delta Lake
 
The Paved Road at Netflix
The Paved Road at NetflixThe Paved Road at Netflix
The Paved Road at Netflix
 
MLOps in action
MLOps in actionMLOps in action
MLOps in action
 
Managing the Machine Learning Lifecycle with MLflow
Managing the Machine Learning Lifecycle with MLflowManaging the Machine Learning Lifecycle with MLflow
Managing the Machine Learning Lifecycle with MLflow
 
Azure DevOps
Azure DevOpsAzure DevOps
Azure DevOps
 
Software estimation
Software estimationSoftware estimation
Software estimation
 
Introduction to GitHub Copilot
Introduction to GitHub CopilotIntroduction to GitHub Copilot
Introduction to GitHub Copilot
 
What Is DevOps?
What Is DevOps?What Is DevOps?
What Is DevOps?
 

Similar to How to deploy machine learning models into production

IBM Bluemix saves the game
IBM Bluemix saves the gameIBM Bluemix saves the game
IBM Bluemix saves the gamegjuljo
 
IBM Bluemix Tech Meetup 18-02-2015
IBM Bluemix Tech Meetup 18-02-2015IBM Bluemix Tech Meetup 18-02-2015
IBM Bluemix Tech Meetup 18-02-2015gjuljo
 
Inteligencia artificial, open source e IBM Call for Code
Inteligencia artificial, open source e IBM Call for CodeInteligencia artificial, open source e IBM Call for Code
Inteligencia artificial, open source e IBM Call for CodeLuciano Resende
 
S104 twist and cloud
S104 twist and cloudS104 twist and cloud
S104 twist and cloudnick_garrod
 
fiu-cloud-hackathon-lec1-v6
fiu-cloud-hackathon-lec1-v6fiu-cloud-hackathon-lec1-v6
fiu-cloud-hackathon-lec1-v6Kirill Osipov
 
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...Creating Microservices Application with IBM Cloud Private (ICP) - introductio...
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...PT Datacomm Diangraha
 
Elevating your Continuous Delivery Strategy Above the Rolling Clouds
Elevating your Continuous Delivery Strategy Above the Rolling CloudsElevating your Continuous Delivery Strategy Above the Rolling Clouds
Elevating your Continuous Delivery Strategy Above the Rolling CloudsMichael Elder
 
How do you deliver your applications to the cloud?
How do you deliver your applications to the cloud?How do you deliver your applications to the cloud?
How do you deliver your applications to the cloud?Michael Elder
 
仕事ではじめる機械学習
仕事ではじめる機械学習仕事ではじめる機械学習
仕事ではじめる機械学習Aki Ariga
 
Accelerate Digital Transformation with IBM Cloud Private
Accelerate Digital Transformation with IBM Cloud PrivateAccelerate Digital Transformation with IBM Cloud Private
Accelerate Digital Transformation with IBM Cloud PrivateMichael Elder
 
Platform as a Service - Cloud Foundry and IBM Bluemix
Platform as a Service - Cloud Foundry and IBM BluemixPlatform as a Service - Cloud Foundry and IBM Bluemix
Platform as a Service - Cloud Foundry and IBM BluemixDavid Currie
 
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...Michael O'Sullivan
 
Http Services in Rust on Containers
Http Services in Rust on ContainersHttp Services in Rust on Containers
Http Services in Rust on ContainersAnton Whalley
 
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018IBM Hybrid Cloud Integration UCC Talk, 21st November 2018
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018Michael O'Sullivan
 
Cloud adoption patterns April 11 2016
Cloud adoption patterns April 11 2016Cloud adoption patterns April 11 2016
Cloud adoption patterns April 11 2016Kyle Brown
 
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015Christophe Lucas
 
IBM Cloud Private and IBM Power Systems: Overview and Real-World Scenarios
IBM Cloud Private and IBM Power Systems: Overview and Real-World ScenariosIBM Cloud Private and IBM Power Systems: Overview and Real-World Scenarios
IBM Cloud Private and IBM Power Systems: Overview and Real-World ScenariosJoe Cropper
 
Bangalore Cloud Foundry meetup - Mani
Bangalore Cloud Foundry meetup - ManiBangalore Cloud Foundry meetup - Mani
Bangalore Cloud Foundry meetup - ManiMani Chandrasekaran
 
Train, predict, serve: How to go into production your machine learning model
Train, predict, serve: How to go into production your machine learning modelTrain, predict, serve: How to go into production your machine learning model
Train, predict, serve: How to go into production your machine learning modelCloudera Japan
 

Similar to How to deploy machine learning models into production (20)

IBM Bluemix saves the game
IBM Bluemix saves the gameIBM Bluemix saves the game
IBM Bluemix saves the game
 
IBM Bluemix Tech Meetup 18-02-2015
IBM Bluemix Tech Meetup 18-02-2015IBM Bluemix Tech Meetup 18-02-2015
IBM Bluemix Tech Meetup 18-02-2015
 
App Modernization
App ModernizationApp Modernization
App Modernization
 
Inteligencia artificial, open source e IBM Call for Code
Inteligencia artificial, open source e IBM Call for CodeInteligencia artificial, open source e IBM Call for Code
Inteligencia artificial, open source e IBM Call for Code
 
S104 twist and cloud
S104 twist and cloudS104 twist and cloud
S104 twist and cloud
 
fiu-cloud-hackathon-lec1-v6
fiu-cloud-hackathon-lec1-v6fiu-cloud-hackathon-lec1-v6
fiu-cloud-hackathon-lec1-v6
 
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...Creating Microservices Application with IBM Cloud Private (ICP) - introductio...
Creating Microservices Application with IBM Cloud Private (ICP) - introductio...
 
Elevating your Continuous Delivery Strategy Above the Rolling Clouds
Elevating your Continuous Delivery Strategy Above the Rolling CloudsElevating your Continuous Delivery Strategy Above the Rolling Clouds
Elevating your Continuous Delivery Strategy Above the Rolling Clouds
 
How do you deliver your applications to the cloud?
How do you deliver your applications to the cloud?How do you deliver your applications to the cloud?
How do you deliver your applications to the cloud?
 
仕事ではじめる機械学習
仕事ではじめる機械学習仕事ではじめる機械学習
仕事ではじめる機械学習
 
Accelerate Digital Transformation with IBM Cloud Private
Accelerate Digital Transformation with IBM Cloud PrivateAccelerate Digital Transformation with IBM Cloud Private
Accelerate Digital Transformation with IBM Cloud Private
 
Platform as a Service - Cloud Foundry and IBM Bluemix
Platform as a Service - Cloud Foundry and IBM BluemixPlatform as a Service - Cloud Foundry and IBM Bluemix
Platform as a Service - Cloud Foundry and IBM Bluemix
 
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...
IBM Hybrid Cloud Integration UCC Talk, 23rd November 2021 - Cloud Application...
 
Http Services in Rust on Containers
Http Services in Rust on ContainersHttp Services in Rust on Containers
Http Services in Rust on Containers
 
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018IBM Hybrid Cloud Integration UCC Talk, 21st November 2018
IBM Hybrid Cloud Integration UCC Talk, 21st November 2018
 
Cloud adoption patterns April 11 2016
Cloud adoption patterns April 11 2016Cloud adoption patterns April 11 2016
Cloud adoption patterns April 11 2016
 
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015
Perth DevOps Meetup - Introducing the IBM Innovation Lab - 12112015
 
IBM Cloud Private and IBM Power Systems: Overview and Real-World Scenarios
IBM Cloud Private and IBM Power Systems: Overview and Real-World ScenariosIBM Cloud Private and IBM Power Systems: Overview and Real-World Scenarios
IBM Cloud Private and IBM Power Systems: Overview and Real-World Scenarios
 
Bangalore Cloud Foundry meetup - Mani
Bangalore Cloud Foundry meetup - ManiBangalore Cloud Foundry meetup - Mani
Bangalore Cloud Foundry meetup - Mani
 
Train, predict, serve: How to go into production your machine learning model
Train, predict, serve: How to go into production your machine learning modelTrain, predict, serve: How to go into production your machine learning model
Train, predict, serve: How to go into production your machine learning model
 

More from DataWorks Summit

Floating on a RAFT: HBase Durability with Apache Ratis
Floating on a RAFT: HBase Durability with Apache RatisFloating on a RAFT: HBase Durability with Apache Ratis
Floating on a RAFT: HBase Durability with Apache RatisDataWorks Summit
 
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFi
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFiTracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFi
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFiDataWorks Summit
 
HBase Tales From the Trenches - Short stories about most common HBase operati...
HBase Tales From the Trenches - Short stories about most common HBase operati...HBase Tales From the Trenches - Short stories about most common HBase operati...
HBase Tales From the Trenches - Short stories about most common HBase operati...DataWorks Summit
 
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...DataWorks Summit
 
Managing the Dewey Decimal System
Managing the Dewey Decimal SystemManaging the Dewey Decimal System
Managing the Dewey Decimal SystemDataWorks Summit
 
Practical NoSQL: Accumulo's dirlist Example
Practical NoSQL: Accumulo's dirlist ExamplePractical NoSQL: Accumulo's dirlist Example
Practical NoSQL: Accumulo's dirlist ExampleDataWorks Summit
 
HBase Global Indexing to support large-scale data ingestion at Uber
HBase Global Indexing to support large-scale data ingestion at UberHBase Global Indexing to support large-scale data ingestion at Uber
HBase Global Indexing to support large-scale data ingestion at UberDataWorks Summit
 
Scaling Cloud-Scale Translytics Workloads with Omid and Phoenix
Scaling Cloud-Scale Translytics Workloads with Omid and PhoenixScaling Cloud-Scale Translytics Workloads with Omid and Phoenix
Scaling Cloud-Scale Translytics Workloads with Omid and PhoenixDataWorks Summit
 
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFi
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFiBuilding the High Speed Cybersecurity Data Pipeline Using Apache NiFi
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFiDataWorks Summit
 
Supporting Apache HBase : Troubleshooting and Supportability Improvements
Supporting Apache HBase : Troubleshooting and Supportability ImprovementsSupporting Apache HBase : Troubleshooting and Supportability Improvements
Supporting Apache HBase : Troubleshooting and Supportability ImprovementsDataWorks Summit
 
Security Framework for Multitenant Architecture
Security Framework for Multitenant ArchitectureSecurity Framework for Multitenant Architecture
Security Framework for Multitenant ArchitectureDataWorks Summit
 
Presto: Optimizing Performance of SQL-on-Anything Engine
Presto: Optimizing Performance of SQL-on-Anything EnginePresto: Optimizing Performance of SQL-on-Anything Engine
Presto: Optimizing Performance of SQL-on-Anything EngineDataWorks Summit
 
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...DataWorks Summit
 
Extending Twitter's Data Platform to Google Cloud
Extending Twitter's Data Platform to Google CloudExtending Twitter's Data Platform to Google Cloud
Extending Twitter's Data Platform to Google CloudDataWorks Summit
 
Event-Driven Messaging and Actions using Apache Flink and Apache NiFi
Event-Driven Messaging and Actions using Apache Flink and Apache NiFiEvent-Driven Messaging and Actions using Apache Flink and Apache NiFi
Event-Driven Messaging and Actions using Apache Flink and Apache NiFiDataWorks Summit
 
Securing Data in Hybrid on-premise and Cloud Environments using Apache Ranger
Securing Data in Hybrid on-premise and Cloud Environments using Apache RangerSecuring Data in Hybrid on-premise and Cloud Environments using Apache Ranger
Securing Data in Hybrid on-premise and Cloud Environments using Apache RangerDataWorks Summit
 
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...DataWorks Summit
 
Computer Vision: Coming to a Store Near You
Computer Vision: Coming to a Store Near YouComputer Vision: Coming to a Store Near You
Computer Vision: Coming to a Store Near YouDataWorks Summit
 
Big Data Genomics: Clustering Billions of DNA Sequences with Apache Spark
Big Data Genomics: Clustering Billions of DNA Sequences with Apache SparkBig Data Genomics: Clustering Billions of DNA Sequences with Apache Spark
Big Data Genomics: Clustering Billions of DNA Sequences with Apache SparkDataWorks Summit
 

More from DataWorks Summit (20)

Data Science Crash Course
Data Science Crash CourseData Science Crash Course
Data Science Crash Course
 
Floating on a RAFT: HBase Durability with Apache Ratis
Floating on a RAFT: HBase Durability with Apache RatisFloating on a RAFT: HBase Durability with Apache Ratis
Floating on a RAFT: HBase Durability with Apache Ratis
 
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFi
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFiTracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFi
Tracking Crime as It Occurs with Apache Phoenix, Apache HBase and Apache NiFi
 
HBase Tales From the Trenches - Short stories about most common HBase operati...
HBase Tales From the Trenches - Short stories about most common HBase operati...HBase Tales From the Trenches - Short stories about most common HBase operati...
HBase Tales From the Trenches - Short stories about most common HBase operati...
 
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...
Optimizing Geospatial Operations with Server-side Programming in HBase and Ac...
 
Managing the Dewey Decimal System
Managing the Dewey Decimal SystemManaging the Dewey Decimal System
Managing the Dewey Decimal System
 
Practical NoSQL: Accumulo's dirlist Example
Practical NoSQL: Accumulo's dirlist ExamplePractical NoSQL: Accumulo's dirlist Example
Practical NoSQL: Accumulo's dirlist Example
 
HBase Global Indexing to support large-scale data ingestion at Uber
HBase Global Indexing to support large-scale data ingestion at UberHBase Global Indexing to support large-scale data ingestion at Uber
HBase Global Indexing to support large-scale data ingestion at Uber
 
Scaling Cloud-Scale Translytics Workloads with Omid and Phoenix
Scaling Cloud-Scale Translytics Workloads with Omid and PhoenixScaling Cloud-Scale Translytics Workloads with Omid and Phoenix
Scaling Cloud-Scale Translytics Workloads with Omid and Phoenix
 
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFi
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFiBuilding the High Speed Cybersecurity Data Pipeline Using Apache NiFi
Building the High Speed Cybersecurity Data Pipeline Using Apache NiFi
 
Supporting Apache HBase : Troubleshooting and Supportability Improvements
Supporting Apache HBase : Troubleshooting and Supportability ImprovementsSupporting Apache HBase : Troubleshooting and Supportability Improvements
Supporting Apache HBase : Troubleshooting and Supportability Improvements
 
Security Framework for Multitenant Architecture
Security Framework for Multitenant ArchitectureSecurity Framework for Multitenant Architecture
Security Framework for Multitenant Architecture
 
Presto: Optimizing Performance of SQL-on-Anything Engine
Presto: Optimizing Performance of SQL-on-Anything EnginePresto: Optimizing Performance of SQL-on-Anything Engine
Presto: Optimizing Performance of SQL-on-Anything Engine
 
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...
Introducing MlFlow: An Open Source Platform for the Machine Learning Lifecycl...
 
Extending Twitter's Data Platform to Google Cloud
Extending Twitter's Data Platform to Google CloudExtending Twitter's Data Platform to Google Cloud
Extending Twitter's Data Platform to Google Cloud
 
Event-Driven Messaging and Actions using Apache Flink and Apache NiFi
Event-Driven Messaging and Actions using Apache Flink and Apache NiFiEvent-Driven Messaging and Actions using Apache Flink and Apache NiFi
Event-Driven Messaging and Actions using Apache Flink and Apache NiFi
 
Securing Data in Hybrid on-premise and Cloud Environments using Apache Ranger
Securing Data in Hybrid on-premise and Cloud Environments using Apache RangerSecuring Data in Hybrid on-premise and Cloud Environments using Apache Ranger
Securing Data in Hybrid on-premise and Cloud Environments using Apache Ranger
 
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...
Big Data Meets NVM: Accelerating Big Data Processing with Non-Volatile Memory...
 
Computer Vision: Coming to a Store Near You
Computer Vision: Coming to a Store Near YouComputer Vision: Coming to a Store Near You
Computer Vision: Coming to a Store Near You
 
Big Data Genomics: Clustering Billions of DNA Sequences with Apache Spark
Big Data Genomics: Clustering Billions of DNA Sequences with Apache SparkBig Data Genomics: Clustering Billions of DNA Sequences with Apache Spark
Big Data Genomics: Clustering Billions of DNA Sequences with Apache Spark
 

Recently uploaded

TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024Lonnie McRorey
 
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfSo einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfpanagenda
 
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptx
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptxUse of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptx
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptxLoriGlavin3
 
Scale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterScale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterMydbops
 
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentEmixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentPim van der Noll
 
Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Farhan Tariq
 
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better StrongerModern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better Strongerpanagenda
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .Alan Dix
 
The State of Passkeys with FIDO Alliance.pptx
The State of Passkeys with FIDO Alliance.pptxThe State of Passkeys with FIDO Alliance.pptx
The State of Passkeys with FIDO Alliance.pptxLoriGlavin3
 
A Deep Dive on Passkeys: FIDO Paris Seminar.pptx
A Deep Dive on Passkeys: FIDO Paris Seminar.pptxA Deep Dive on Passkeys: FIDO Paris Seminar.pptx
A Deep Dive on Passkeys: FIDO Paris Seminar.pptxLoriGlavin3
 
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...Wes McKinney
 
Sample pptx for embedding into website for demo
Sample pptx for embedding into website for demoSample pptx for embedding into website for demo
Sample pptx for embedding into website for demoHarshalMandlekar2
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfMounikaPolabathina
 
How to write a Business Continuity Plan
How to write a Business Continuity PlanHow to write a Business Continuity Plan
How to write a Business Continuity PlanDatabarracks
 
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxMerck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxLoriGlavin3
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityIES VE
 
A Framework for Development in the AI Age
A Framework for Development in the AI AgeA Framework for Development in the AI Age
A Framework for Development in the AI AgeCprime
 
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...Alkin Tezuysal
 
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxPasskey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxLoriGlavin3
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxLoriGlavin3
 

Recently uploaded (20)

TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024
 
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfSo einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
 
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptx
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptxUse of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptx
Use of FIDO in the Payments and Identity Landscape: FIDO Paris Seminar.pptx
 
Scale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterScale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL Router
 
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentEmixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
 
Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...
 
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better StrongerModern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .
 
The State of Passkeys with FIDO Alliance.pptx
The State of Passkeys with FIDO Alliance.pptxThe State of Passkeys with FIDO Alliance.pptx
The State of Passkeys with FIDO Alliance.pptx
 
A Deep Dive on Passkeys: FIDO Paris Seminar.pptx
A Deep Dive on Passkeys: FIDO Paris Seminar.pptxA Deep Dive on Passkeys: FIDO Paris Seminar.pptx
A Deep Dive on Passkeys: FIDO Paris Seminar.pptx
 
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
 
Sample pptx for embedding into website for demo
Sample pptx for embedding into website for demoSample pptx for embedding into website for demo
Sample pptx for embedding into website for demo
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdf
 
How to write a Business Continuity Plan
How to write a Business Continuity PlanHow to write a Business Continuity Plan
How to write a Business Continuity Plan
 
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptxMerck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a reality
 
A Framework for Development in the AI Age
A Framework for Development in the AI AgeA Framework for Development in the AI Age
A Framework for Development in the AI Age
 
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
Unleashing Real-time Insights with ClickHouse_ Navigating the Landscape in 20...
 
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxPasskey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
 

How to deploy machine learning models into production

  • 1. IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation 1 How to deploy machine learning models into production Sumit Goyal (@SumitG0yal) Software Engineer, Watson Studio Sven Hafeneger Data Scientist, Watson Studio
  • 2. IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation 2 About me Sumit Goyal (@SumitG0yal ) Software Engineer, Watson Studio
  • 3. Agenda 3IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 4. Motivation 4IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining Ops Dev
  • 5. Motivation 5IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining Work does not end with handover of ML models !
  • 6. Motivation 6IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation http://pobsonline.com/money-mistakes/, https://www.drwf.org.uk/news-and-events/events/100-day-challenge Mike Data Scientist Deb The Developer ModelDev Ops Production Environment
  • 7. Motivation 7IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation ,http://pobsonline.com/money-mistakes/, https://www.drwf.org.uk/news-and-events/events/100-day-challenge , https://de.wikipedia.org/wiki/Zahnrad Mike Data Scientist Deb The Developer Model + App DevOps Production Environment
  • 8. Motivation 8IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cross-industry_standard_process_for_data_mining • Production environment is in the Cloud • Python as primary programming language • Micro-service architecture • Python ML Frameworks • Traditional ML algorithms, no deep learning • Data Science Team <-> Software Development Team
  • 9. Agenda 9IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 10. ML Model artifacts 10IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation PMML Serialized Object Re-Implementation in Java/C++ Python ML Model Production Environment App REST API App
  • 11. ML Model artifacts 11IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation PMML Serialized Object Re-Implementation in Java/C++ App REST API App Python ML Model Production Environment Handover model, code, integration test, no re-train in Travis
  • 12. Agenda 12IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 13. Production Environments 13IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Serialized Object (Python) Meta data - Version - Name - Type - ML Framework - Eval. Metrics - … Python Application - REST API Endpoint: /v1/xyz/prediction : POST Webserver / Web framework Endpoint ML framework - Load ML model - Create a route - Prepare data - Run prediction - Post-processing - Return result Logic https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html
  • 14. Production Environments – eCommerce 14IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Endpoint REST API App Service agreement - max. response time - availability - quality/confidence of prediction - max. re-train time - monitor - … https://blog.paymill.com/en/keep-e-commerce-system-functioning-optimum-capacity/, http://www.atlascorps.org/blog/customers-and-clients-the-devils-of-business-growth/ App App
  • 15. Production Environments – eCommerce 15IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Endpoint REST API App Service agreement - max. response time - availability - quality/confidence of prediction - max. re-train time - monitor - … https://blog.paymill.com/en/keep-e-commerce-system-functioning-optimum-capacity/, http://www.atlascorps.org/blog/customers-and-clients-the-devils-of-business-growth/ App App Order == Fraud ? Order!
  • 16. Production Environments – Fraud Detection 16IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://www.kaggle.com/mlg-ulb/creditcardfraud Ordering Transactions
  • 17. Agenda 17IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 18. Deployment with Cloud Foundry 18IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Python App - REST API Endpoint: /v1/xyz/prediction : POST REST - Load ML model - Create a route - Prep. score data - Run prediction - Post-processing - Return result Model https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org Questions - How to configure the app ? - How many instances ? - How much RAM ? - Zero downtime deployment ? - Connection to DB? - Logging ? - Crashes and recovery? - Routing ? - Workload scaling ? - Integration with other tools? - …
  • 19. Deployment with Cloud Foundry 19IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Choose a build pack !
  • 20. Deployment with Cloud Foundry 20IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Python App: /route/.. http://flask.pocoo.org/, http://scikit-learn.org ../projects/fraud_detection_api CLI: push CLI: set-env Configuration R E S T A P I Code App POST: {body} Order == Fraud ?
  • 21. Deployment with Cloud Foundry: Demo 21IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Cloud_Foundry
  • 22. Agenda 22IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 23. Deployment with Kubernetes 23IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Container - REST API Endpoint: /v1/xyz/prediction : POST Endpoint - Create a route - Load ML model - Prepare data - Run prediction - Post-processing - Return result Logic https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org Questions - How to configure the app ? - How many instances ? - How much RAM ? - How many CPUs/GPUs? - Zero downtime deployment? - Connection to DB? - Logging ? - Crashes and recovery? - Routing ? - Workload scaling ? - Integration with other tools? - … - How much do you want to customize the runtime?
  • 24. Deployment with Kubernetes 24IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation Docker Container REST API Logic Docker Images Repository App (Container) POST: {body} https://www.iconsdb.com/caribbean-blue-icons/database-5-icon.html , http://flask.pocoo.org/, http://scikit-learn.org, https://blog.takipi.com/pivotal- cloud-foundry-vs-kubernetes-choosing-the-right-cloud-native-application-deployment-platform/ Order == Fraud ?
  • 25. 25IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://en.wikipedia.org/wiki/Kubernetes Deployment with Kubernetes: Demo
  • 26. Agenda 26IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivation • Python ML Model artifacts • Production Environments • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 27. Deployment with managed services 27IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation ML Models Deployment R E S T A P I App POST: {body} Order == Fraud ? Manages Deployment Service
  • 28. Deployment with managed services 28IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://docs.microsoft.com/en-us/azure/machine-learning/, https://aws.amazon.com/sagemaker/, https://console.bluemix.net/catalog/services/machine-learning IBM Watson Machine Learning
  • 29. Deployment with managed services: Demo 29IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation https://console.bluemix.net/catalog/services/machine-learning
  • 30. Summary 30IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Motivated and informed • Deployment with Cloud Foundry • Deployment with Kubernetes • Deployment with managed services
  • 31. Learn more about: 31IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation • Cloud Foundry: https://www.ibm.com/cloud/cloud-foundry • Kubernetes: https://kubernetes.io/ • Watson Machine Learning: https://www.ibm.com/cloud/machine-learning
  • 32. Thank you for your attention. Time for QA! Sumit Goyal sumit.goyal@de.ibm.com 32IBM Cloud / Watson and Cloud Platform / © 2018 IBM Corporation