our portfolio how to build a modern data analytics · fred is the co-founder and cto of aptitive, a...
TRANSCRIPT
Our PortfolioInsert subtitle right here
Fred Bliss, CTO, Aptitive
How to Build a Modern Data
Analytics Platform
Fred is the co-founder and CTO of Aptitive, a data and analytics
consulting firm in Chicago. We work primarily with midsize to large
enterprise organizations to develop custom applications, data
pipelines, and analytical data platforms - all to help make data
accessible and meaningful to systems and people who need to
make sense of it. By taking a tech/vendor agnostic approach, we
help implement and architect these solutions using the best
technology and tools that fit an organization’s needs.
Fred lives in Chicago with his wife, Julia, and his daughters, Claire
and Zoe. He spends way too much time playing with and reading
about tech, and always enjoys getting in front of a whiteboard
and collaborating on a solution with a group of open-minded
visionaries.Fred Bliss, CTO, Aptitive
A Short History of AnalyticsThe History of Decision Support Systems, Business Intelligence, and Analytics
1980 - 2000
Ad-hoc queries against operational
systems, formatted reports
Rise of the enterprise data
warehouse and ‘self-service’
2008 - 2015
Rise of dashboards and visual
analytics, “Big Data” era
Cloud, Machine Learning, faster
speed-to-insight for business -
enter the modern data platform
Early 2000’s Current Day
In the first 10-15 years of the 2000’s, the technical design of Business Intelligence projects began to resemble that of an
assembly line in a factory.
A Short History of Analytics
Dashboard tools sold at the department level resulted in
data silos across organizations, but also increased the hunger
for visual analytics.
A Short History of Analytics
The “Big Data” era of Hadoop was largely a flop in enterprise
organizations, but it planted the seeds for the innovative ecosystem
we have today.
A Short History of Analytics
Today, in the modern data world, the complexity is not in the tooling - it’s in navigating the choices of which ones to
use, and how they work together.
A Short History of Analytics
Our WorksInsert subtitle right here
The Impact on Data Professionals
Whether you’re just starting your career or a
seasoned data professional, the path to
success is the same - continuous learning.
Meet Our TeamInsert subtitle right here
The Impact on Data Professionals
Yesteryear, data professionals could get by knowing a handful of technology stacks and design patternsData Professionals’ Experience
Business Intelligence Ecosystem
Meet Our TeamInsert subtitle right here
The Impact on Data Professionals
Today, it’s impossible to have experience in everything. What does this mean for you and for hiring managers?Data Professionals’ Experience
Modern Data Platform Ecosystem
Table of ContentInsert subtitle right here
01.Step Ask ‘Why?’
02.Step 2Ask ‘Why?’ Again
04.Step 4Ask ‘Why?’ Again
05.Step 5Ask ‘Why?’ Again
03.Step 3Ask ‘Why?’ Again
06.Step 6Ask ‘Why?’ Again
Prototype. Build. Experience.The Path to Success
Be Curious About Tech.
"Your brilliant first flop was a raging success! Come on, let’s get busy and on to the next!"
She handed a notebook to Rosie Revere, who smiled at her aunt as it all became clear.
Life might have its failures, but this was not it. The only true failure can come if you quit.”
Source: https://www.scholastic.com/parents/books-and-reading/raise-a-reader-blog/rosie-revere-engineer-lesson.html
Our WorksInsert subtitle right here“No matter where you are in your analytics journey, the key to
long-term success is a modern data platform”
“The elements of a successful, modern data platform are much more than dashboards, reports,
and predictive insights”
What is a Modern Data Platform ?
Our Lambda Design Reference Architecture
On-PremApplications
SaaS Applications
Business Applications
S3/ Blob Storage Cloud Data
Platform
Machine Learning Models
Batch Replication Path
Write Back Results from
Model
Real-timeEvent
Processing
Da
ta W
are
hous
e
Raw Tables
Business Tables
Virtualized Flattened ML
Schema
Virtualized Star Schema
Internal Analytics
Embedded Customer Analytics
Semantic Model
Ana
lytic
s La
yer
What is a Modern Data Platform ?
When Building a Modern Data
Platform, You Can Break Up the
Technical Components Into
the Following:Advanced AnalyticsVisualization
Orchestration
TransformationIngestion
On-PremApplications
SaaS Applications
Business Applications
S3/ Blob Storage
Batch Replication Path
Real-timeEvent
Processing
Ingestion
“How do we get our data?”
Transformation
“What is the business logic and data model required to integrate
and present our data?”
Cloud Data Platform
Machine Learning Models
Write Back Results from
Model
Da
ta W
are
hous
e
Raw Tables
Business Tables
Virtualized Flattened ML
Schema
Virtualized Star Schema
Internal Analytics
Embedded Customer Analytics
Semantic Model
Ana
lytic
s La
yer
Orchestration
“How do we automate the process?”
Visualization
“How do we present the data for analytical
insights?”Internal
Analytics
Embedded Customer Analytics
Semantic Model
Ana
lytic
s La
yer
Analytics & BI Tools
Advanced Analytics
“How can we use our data for both
predictive capabilities and automated
discovery of insights?”Cloud Data
Platform
Machine Learning Models
Write Back Results from
Model
Prediction APIs
Data Science Toolbox
Consumer Apps
01.The End Product & Vision
02.Domain Knowledge
04.Tools vs. Knowledge
05.Unproven Technology
03.The Right Approach
06.Over-Engineering
“What are some pitfalls to consider and things to watch out for?”
Getting Started with a Modern Data Platform
Guiding Principles❏ Avoid over-dependence on a single application
❏ Speed to Value
❏ Easily Maintainable
❏ Write Code Once Mentality
❏ Consistent Business Logic and Data Definitions
❏ Start Small, Think Big
❏ Plan for Future Change and Disruption
Typical Current State Challenges
❏ Multiple sources of truth for data
❏ Business logic buried in BI tools or Excel
❏ Point-to-point mentality of integration flows across the enterprise
❏ Inflexible data and analytics
❏ Security is considered later
❏ Narrow use cases that define the broader architectureAnalytics tools can create new data silos
Tips on Preparing for the Future
Get Started Today with a Modern Data Architecture Whiteboard Session.
www.aptitive.com/snowflake
Crystalize the design and architecture of your
future state data and analytics environment.
www.aptitive.com/modern-data-architecture-quickstart/
Contact UsInsert subtitle right hereAptitive is a Chicago-based IT
consulting firm with solutions focusing
on the management, design, and
development of data-centric projects.
Thank You!
20 N Wacker Dr, Chicago, IL
www.aptitive.com
312.725.8553
We’re Hiring!