keynote: the anaconda roadmap | anacondacon 2017
TRANSCRIPT
PowerPoint Presentation
#OpenDataScienceMeans#AnacondaCON
Anaconda RoadmapThe Journey to Open Data Science Begins with Anaconda
How is your day going? I hope youve learned lots here at AnacondaCON today!
Id like to introduce my colleagues & your Product Management team for Anaconda & Anaconda Enterprise.
Crystal Soja, Christine Doig, and Kris Overholt!
At the end, well have a Q&A session and they will be assisting me. Id also encourage you to connect with them while youre here and let us know whats important to your enterprise.
2
The Beginning
#OpenDataScienceMeans#AnacondaCON
Id like to start out our journey with two men that are well known to most of you, Travis Oliphant and Peter Wang, our co-founders. Five years ago these two icons in the open source Python community who were known for their foundational contributions to both scientific computing and visualization came together to found Continuum because they shared a common understanding that there were issues preventing their beloved Python from reaching its full potential.
(click) As loyal Pythonistas, they wanted to see Python realize its fullest potential. Yet the anarchy and chaos of packages was clearly preventing wide scale adoption of Python. So they put their big brains together, and yes, thats a pretty powerful combination, to solve just that problem. Out of that collaboration and humble beginnings emerged conda, a cross platform, multi-language package manager, which makes it easy to manage not only Python, but also R, Java, C/C++, Scala and even Fortran.
That was the beginning of their vision for open data science which evolved into Anaconda. And yes, thats the original logo. As you can see, since then, weve come a long way baby and were not stopping. We, with their leadership and technical expertise, are continuously reevaluating the open source ecosystem to help evolve open data science so you dont have to.
3
Youre in good company11 Million Downloads & 229% YoY growth
#OpenDataScienceMeans#AnacondaCON
In 2016, Anaconda achieved widespread market adoption. We went from 3M total downloads at the end of 2015 to 11M total downloads in 2016.
So rest assured, youre in great company!
4
is the leading Open Data Science platform powered by Python the fastest growing data science languageAccelerate Time-to-ValueConnect Data, Analytics & ComputeEmpower Data Science Teams
#OpenDataScienceMeans#AnacondaCON
Anaconda is the leading Open Data Science platform powered by Python and has become the de facto open data science platform of enterprises, scientists, & academics.5
March 2015
February 2016
Feb & Sept 2016
Sept 2016
#OpenDataScienceMeans#AnacondaCON
As well as technology giants worldwide
6
INNOVATE faster through managed agile experimentation MOVE from analysis to deployment immediatelyDELIVER powerful results backed by high performance open data science platform
LEVERAGE innovative open source analytics to extract value from dataMAXIMIZE your computational power to easily analyze all dataCONNECT and integrate all your data sources for predictive models
ITERATE quickly to create powerful analysis and predictive modelsCOLLABORATE and share with your data science teamPUBLISH interactive results to the business
ACCELERATETime-to-ValueCONNECTData, Analytics & ComputeEMPOWERData Science Teams
#OpenDataScienceMeans#AnacondaCON
Because Anaconda
ACCELERATES time-to-valueCONNECTS data, analytics & computeEMPOWERS data science teams7
Gives Superpowers To People Who Change The World
#OpenDataScienceMeans#AnacondaCON
8
#OpenDataScienceMeans#AnacondaCON
As we continue down this journey to Open Data Science weve considered many different paths.
9
#OpenDataScienceMeans#AnacondaCON
Weve gone to the Big Data mountain to seek wisdom10
#OpenDataScienceMeans#AnacondaCON
And weve gone to the promised land of analytics to seek enlightenment
11
#OpenDataScienceMeans#AnacondaCON
and weve looked at the magic of automated machine learning12
#OpenDataScienceMeans#AnacondaCON
And we have taken into account the most demanding requests of enterprise operations
13
DATASCIENCECOLLAB
OPENDATASCIENCE
DATASCIENCEGOVERNANCE
DASHBOARDS& APPS
SELFSERVICEANALYTICS
DATASCIENCEOPERATIONS
DATASCIENCE FORBIG DATA
AI
#OpenDataScienceMeans#AnacondaCON
To continue pushing the boundaries towards an enlightened open data science path a path where we embrace the best that open source has to offer, while we continue to fill the gaps in the ecosystem for enterprises by innovating and contributing foundational technology to the open data science ecosystem.
While many are now claiming to be a data science platform, they dont fully embrace the promise of open data science to help you connect the open source ecosystems to enterprise requirements. While we understand many of the complex enterprise requirements we are always open to learning more from you about what you need to unleash the value locked up in your data. Thats why Anaconda Enterprise is focused on empowering your data science teams with (click)
Because thats what it takes to be an enterprise ready open data science platform.
14
Empower the data science team
Empower Data Science teams to
#OpenDataScienceMeans#AnacondaCON
With those capabilities your data science team is empowered to Explore & analyze their dataCollaborate & publish their data science modelsDeploy & operate their data science applications to realize value for your enterprise
15
Lego Movie
#OpenDataScienceMeans#AnacondaCON
Lets take a peek at a data science team in action.
If you havent found your data science team yet, theyre awaiting you at the Anaconda demo kiosks!16
#OpenDataScienceMeans#AnacondaCON
While we know quite a bit about open data science, our inquiring minds still wanted to know more about what you, the enterprise innovators and leaders, need from open data science17
First Open Data Science Survey
#OpenDataScienceMeans#AnacondaCON
So we commissioned the first ever open data science survey that will be sent to each of you after the conference18
97% enterprises report Data Science is critical to the success of their business
#OpenDataScienceMeans#AnacondaCON
And we discovered that .19
94% enterprises using open source for Data Scienceand more than half (54%) report that their technology philosophy for data science is entirely or mostly open source
#OpenDataScienceMeans#AnacondaCON
And enterprises are relying on open source for enterprise data science20
73% enterprises report Data Science is in Top 3 technologies that bring the most value to their organization
#OpenDataScienceMeans#AnacondaCON
And that data science is one of the most import technologies that is allowing companies to get value from their data especially their Big Data21
Open Data Science means Collaboration
#OpenDataScienceMeans#AnacondaCON
And that OpenDataScience means effective team COLLABORATION 22
86% of enterprises are actively using data science in their business
Bottom line is that data science is no longer just for competitive advantage its infused into day-to-day operations.
Data science is business.
#OpenDataScienceMeans#AnacondaCON
And that in this new world order, data science is actively being used by enterprises not just for competitive advantage but to run the business and drive the valuation gains that Brian Hopkins talked about earlier today. We learned that Data Science is business. And that the best run businesses run on open data science.
23
Anaconda RoadmapIntroduction toSAFE HARBORCertain information contained in this presentation is forward-looking in nature. Any expectations based on these forward-looking statements are subject to risks and uncertainties and other important factors. These and many other factors could cause delivery of products, features or enhancements to differ materially from expectations based on these forward-looking statements. Continuum Analytics does not undertake an obligation to update its forward-looking statements to reflect future events or circumstances.
#OpenDataScienceMeans#AnacondaCON
We took all of this information plus our open data science expertise to forge a roadmap to help our customers realize their ambitious plans to change the world
24
Anaconda today
Open Data Science CoreHigh Performance ComputingData Science CollaborationExcel Data ScienceDistributed ComputingOpen Data Science HubANACONDAANACONDA REPOSITORYANACONDA ACCELERATEANACONDA SCALEANACONDA ENTERPRISE NOTEBOOKSANACONDA FUSION
#OpenDataScienceMeans#AnacondaCON
We looked at where we are today with Anaconda being a full-stack, open data science platform with many components.
By the way, I should mention that today, Crystal manages the Anaconda and Anaconda Repository teams. Kris manages the Adam, Accelerate, Scale and Enterprise Notebook teams. Christine manages the Fusion team. 25
Anaconda On-Prem First, Cloud Ready
ON-PREMISEPRIVATE CLOUDANACONDA CLOUD
#OpenDataScienceMeans#AnacondaCON
And we remain commited that Anaconda will be on-premise first AND cloud-ready 26
AnacondaHigh performance Python & R720+ data science packagesCross-platform package, dependency & environmentsCommunity driven package repository collaboration Anaconda NavigatorDesktop Portal & InstallerAnaconda PowersOPEN DATA SCIENCEDATA SCIENCE GOVERNANCEDATA SCIENCE COLLABORATIONAnaconda RepositoryStorage & sharing of packages, environments, notebooksOn-premise governanceEnterprise authenticationAnaconda Enterprise NotebooksCollaborative project based workflows for Python & REnterprise authentication & permissioningNotebook sharing, versioning, search, differencingDATA SCIENCE FOR BIG DATAAnaconda Scale Hadoop & Spark integrationScalable distributed processing frameworkIntegration with resource management & data storesSelf-service cluster launchingDistributed package, dependency & environments Anaconda FusionBig Data querying & transformations
#OpenDataScienceMeans#AnacondaCON
#
And that well continue to deliver on the breadth of capabilities required of a full end-to-end platform that you, our enterprise customers need and depend upon for your business.
27
Michele Chambers (MC) - Anaconda720+ data science packagesDeep Learning: Theano, Tensorflow, Caffe, Keras, Neon, LasagneNatural Language Processing: NLTK, spaCyMachine Learning: Scikit-learnGPU enablement
Anaconda PowersAIDASHBOARDS & APPSDATA SCIENCE OPERATIONSAnacondaInteractive browser based dashboards & visualizations with BokehBokeh apps using Python, R, ScalaBig Data visualizations with DataShaderAnaconda AdamServer & Cluster InstallerAnaconda AcceleratePython compilation for multi-core & GPUsCode, data, in-notebook profilersPre-optimized numerical librariesSELF-SERVICE ANALYTICSAnaconda FusionIntegration of Open Data Science with Microsoft ExcelInteractive exploration & visualizationPredictive modelingBig Data querying & transformations
#OpenDataScienceMeans#AnacondaCON
#
Yet we know we need to continue evolving Anaconda Enterprise and we have. 28
ANACONDAFUSIONANACONDA ENTERPRISE
ANACONDA
ANACONDANARRATIVES
#OpenDataScienceMeans#AnacondaCON
Today, were introducing to you the next generation Anaconda Enterprise platform with a new unified and streamlined environment to make it faster and easier than ever for data science teams to build innovative, powerful intelligent applications.
TRANSITION TO KRIS
Kris, would you mind giving us a tour of the new Anaconda Enterprise 5.0 platform? 29
Deploy interactive notebooks & applications
DATA SCIENCE EXPERIENCEDATA SCIENCE DEPLOYMENTEnd-to-end collaborative Open Data Science workflowsEncapsulate and share data science assets & applicationsExecute and query machine learning models via REST APIsManage & share data science projects & dependencies
#OpenDataScienceMeans#AnacondaCON
30
#OpenDataScienceMeans#AnacondaCON
31
Demo Movie
#OpenDataScienceMeans#AnacondaCON
ENSURES AVAILABILITY, UPTIME, & MONITORING
PROVISIONS COMPUTE RESOURCES
MANAGES DEPENDENCIES & ENVIRONMENTS
SHARE COMPUTE RESOURCES
SECURE NETWORK COMMUNICATIONS & SSL
SECURE DATA & NETWORK CONNECTIVITY
ENGINEER FOR SCALABILITY
MANAGE AUTHENTICATION & ACCESS CONTROL
SCHEDULE REGULAR EXECUTION OF JOBSWith Anaconda Enterprise life just got a whole lot easierLearn more: https://www.continuum.io/blog/developer-blog/productionizing-deploying-data-science-projects
#OpenDataScienceMeans#AnacondaCON
33
Start with next gen Anaconda Enterprise now
FEBAPRMAYJUNSEPT5.0 Public Cloud Sandbox
Experience end-to-end platform Example data science deployment workflows5.0 Private Cloud Sandbox
Upload data science projects from desktop for deployment Connect to existing Anaconda Repository 4.x5.0 On Premise Sandbox
Upload data science projects from desktop for deployment Connect to existing Anaconda Repository 4.x
Innovator ProgramEarly Adopter ProgramGeneral Availability5.0 On Premise Sandbox
Edit and share notebooks and projects with collaborative workflows5.0 On Premise Production
Mirror Anaconda packages and channels Replicate package repository on premises
#OpenDataScienceMeans#AnacondaCON
5.0 Cloud SandboxExample data science workflows includingRunnable notebooksInteractive data science appsML models with Rest API34
#OpenDataScienceMeans#AnacondaCONTake advantage of the next generation ANACONDA ENTERPRISE now
Join Innovator Program today
Apply: http://go.continuum.io/anaconda-enterprise-innovator/
#OpenDataScienceMeans#AnacondaCON
TRANSITION TO MICHELE
Michele, can you tell us more about what to expect in 2017? 35
Q&A
#OpenDataScienceMeans#AnacondaCON
Weve got a few minutes if anyone would like to ask any questions, weve got a few microphones that well bring to you so everyone can hear your question.
If you have any questions or feedback, please dont hesitate to let us know tonight, tomorrow or at anytime after the conference. 36
Open Data Science Innovation Award
#OpenDataScienceMeans#AnacondaCON
As hopefully youve seen and heard today, many of you are innovators doing amazing breakthrough work in your field.
We wanted to recognize one customer in particular who has been an innovator since their inception almost 30 years ago. This business runs on analytics. It is why they carved out a lucrative market and have become a giant in their industry. Today, they are reinvigorating their business and pushing the boundaries with Open Data Science innovation. (click) Wed like to recognize Hussain Sultan and the team from Capital One for their innovation. Hussain, can you join me? 37
Special AWARDS
BIGGEST DATA SCIENCE TEAM
UNSUNG HEROES
LONGEST DISTANCE
#OpenDataScienceMeans#AnacondaCON
Biggest Data Science Team (12 total)USG (6) Krissy Freeman, Jarek Sychtysz, Rebecca Ward, Patrick Carlos, Alexander Basil, Scott Stevenson State Farm (6) Bob Cunningham, Carlee Clymer, Taylor Smith, Jason White, Peter Laube, Sandra TuckerLongest Distance (9 total)Niko Ahonen and Paavo Pere both from Helsinki Daniel Grafstrom from Sweeden Jens Nie and Peer Wagner from Germany Julien Meltz, Philippe Trin, Haithem Derbel, Julien Lafaye from FranceUnsung Heroes Government (7 total)Vesta Gueschkova, David Lyle, Alexander Gude, Matthew Bement, Andrew Fraser, Dharhas Pothina, Charles CollverMore Unsung Heroes - City & State that we hope will save us! (5 total)City of Boston - Andrew Therriault, Christopher Dwelley, Sam Lovison, Kayla Patel, Alex Chen38
Download your favorite Anaconda wallpaper at:https://www.continuum.io/anaconda-wallpapers
#OpenDataScienceMeans#AnacondaCON
Whos ready to party? 39
#OpenDataScienceMeans#AnacondaCON