Time to Fly - Why Predictive Analytics is Going Mainstream

Welcome

Host:
Eric Kavanagh
eric.kavanagh@bloorgroup.com

Twitter Tag: #briefr The Briefing Room

Mission

!   Reveal the essential characteristics of enterprise
software, good and bad

!   Provide a forum for detailed analysis of today s
innovative technologies

!   Give vendors a chance to explain their product to
savvy analysts

!   Allow audience members to pose serious questions...
and get answers!


November: Cloud

December: Innovators

January: Big Data

February: Analytics


Cloud

!   Cloud computing has come a long way and can now offer an
array of hosted services: software (SaaS), platform (PaaS),
infrastructure (IaaS), as well as other services and resources
(storage, security, API desktop, etc.).

!   Cloud services can be deployed as public, private, or as a
hybrid according to needs.

!   Using the Cloud means companies can invest less in
hardware resources, which can close the gap between large
and mid-sized organizations.


Analyst: Robin Bloor

Robin Bloor is
Chief Analyst at
The Bloor Group

robin.bloor@bloorgroup.com


Perceivant

! Perceivant offers a SaaS-based Big Data analytics platform.

!   The Platform delivers real-time access to disparate data
sources and allows customers to leverage high performance
analytics on terabytes of data.

! Perceivant designed its platform for scalability and ease of
use, requiring no servers, software licenses or consultants.

! Perceivant lets customers begin with a low-risk, short term
license which can be extended after adoption.


Brian Rowe

Brian Rowe is CEO at Perceivant. He has been an
entrepreneur for the last 10 years and started
developing Internet based applications in 1990. He
has been Project Manager for various programs with
annual budgets between $10m-$20m. As Director of e-
Business at Cummins Engine Company, Brian started
six online businesses and won CIO Magazine's 50 Best
Intranets Award during his tenure. Brian had P&L
responsibility at SourceAlliance.com to develop a
Distributed Storefront online retail offering. He was
the COO and co-founder of Redtrain, a Software
Service consultancy. Brian was also a founder of Last
Piece Software where he acted as President until the
company was acquired by iGoDigital. With a
background that bridges both business and
technology, he enjoys the fast paced and dynamic
startup environment, where these skills can be used
effectively. He has a B.S. in Financial Planning and an
MBA from Purdue University.


Background

● Founded in 2012 based on experience working
with data at iGoDigital
● Leveraging nearly a decade of big data and
analysis experience acquired while scaling the
recommendations business at iGoDigital for
some of the biggest names in online retail
● Some of the issues iGoDigital had to solve
became desirable to offer as a stand alone tools

How we got here

● We had 5 TB of data to analyze
● Receiving data in all different formats from
many sources
● Data had to be prepared for analysis - Joined,
flattened, cleansed
● Machine learning algorithms had to run
● Additional needs to manually analyze and
report on data
● Cash constrained as a startup

Searching for a solution

● So many vendors, so little time
● Lots of vendors had partial solutions
● Total cost to own and operate the solutions
were far outside our budget 250K/year and up.
● So we started building....
○ Our goal was the fastest path to ROI

○ Hadoop, Hive, ElasticSearch, MongoDB....

● Started to find out that other people had our
same challenges

A lot of work to do

● Data Preparation - Clean & Flatten
● Visualize & Analyze
● Make a hypothesis, What to predict
● Analysis
● Implement result
● Back Test
● Integrate into business process
● Operate it with scaling compute resources
● Measure it

Big Data issues for businesses

● McKinsey expects that 60% of big data projects
will face a gap in talent by 2018
● People with experience in Hadoop, Hive, Pig,
ElasticSearch, NoSQL, MongoDB, R, SAS, etc.
● Data preparation time typically takes 40-50%
of an analytics project (ETL/Cleaning)
● TDWI study says data prep is taking longer in
2011 than it was in 2009
● Internal data growing 40% annually (IDC
estimate), social data explosion

Breakthrough Insights

● Target predicts fairly well if a shopper is
pregnant based on buying habits and markets
to her upcoming purchasing needs
● Combining credit scores with channel data to
better price loans increasing channel
profitability
● Insights like these can have a big impact on
your business and profitability
● You have already made sizeable investments in
your data, get value from it!

How Perceivant Can Help
● Launch within 30 days
● Achieve 50% savings compared to
purchase package solutions
● BI Suite of tools - Reporting, OLAP,
charting
● Data manipulation - ETL, Cleansing
● Partner with you on predictive
engagements
● Access data anywhere you have the
internet
● No need to worry about scaling
● Easy to integrate back into your existing
systems and process

Predictive Analytics

● Predictive Analytics is not "one size fits all"
● These are services engagements
● We help determine the right combination of
human expertise and machine learning tools
● We deliver and implement the predictive
formula
● We measure the on-going results
● This approach identifies hidden
patterns in your data that will not be
found with traditional approaches

Case Study

● Educational institution spending about 500K/year on
current BI tools and infrastructure, not including people
● Limited capabilities, slow, no predictive capabilities
● RFP produced projects starting at $1M from well known
vendors
● Our solution loaded data in 2 weeks
○ BI tools out of the box satisfied many reporting needs
○ Real-time query capabilities previously not possible
○ Conversion of remaining existing reports
○ Creation of new reports including predictive reports
○ Creation of history from ongoing point in time data
○ Reduced cost by 50% from the current solution

Invest in ROI, not R&D

● The fastest path to ROI on the market today
○ Less than 30 days on-boarding process

○ Monthly starting at $2,000

○ Leverage sizeable investments already made

in your data

Time to Fly - Why Predictive Analytics is Going Mainstream

Perceptions & Questions

Analyst:
Robin Bloor


A NEW
UNIVERSE
OF BI?

The Bloor Group

Open Source, The Cloud and BI (1)

•  Open Source and The Cloud walk
hand in hand
•  Together they undermine the
traditional DBMS business model
in some areas
•  Hadoop and Hadoop++ can happily
be deployed in the cloud
•  Hadoop has become the staging
place for data and an ETL engine
•  The cloud is a natural prototyping
environment - a low risk and
possibly permanent location for BI
apps

The Bloor Group

Open Source, The Cloud and BI (2)

•  The public cloud is growing at >100% p.a.

•  The private cloud is a gateway drug to the public cloud – and vice versa

•  The cloud devalues many software brands (you can only brand applications
now)

•  Open source is beginning to dominate BI cloud deployments by virtue of
cost

•  For BI users, the Cloud + Open Source is
–  Good enough or better
–  A leap-frog opportunity

The Bloor Group

BI Categories

Hindsight: Regular reporting/operational BI
Oversight: Dashboards, OLAP, BPM, etc.
Insight: Data mining, statistical analysis (trends and relationships)
Foresight: Predictive analytics

The Bloor Group

The New BI Universe (?)

The Bloor Group

!  Roughly how much of the software that Perceivant
deploys is home-grown in terms of distinct
components, i.e., what do you add to the solution?

!  Which other data stores/DBMSs do you use, aside
from Hadoop? Any favorites?

!  Is there any aspect of BI that you don’t or wont
cater for (CEP, data governance, MDM, etc.)?

!  What kind of predictive capabilities do you
provide? Are these standard or bespoke?

The Bloor Group

!  Aside from the customer base is there any aspect
of your software stack that could not be
implemented at higher scale (e.g., for large
companies)?

!  How do you define requirements?
!  How many of your customers take the data
scientist consultancy?

!  Roughly how many of your customers eventually
choose to specialize the solution?

The Bloor Group

Upcoming Topics

This month: Cloud

December: Innovators

January: Big Data

2013 Editorial Calendar
www.insideanalysis.com


Thank You
for Your
Attention
Image credits:
http://www.123rf.com/photo_11464614_cloud-glossy-icon-vector-illustration.html
http://www.123rf.com/photo_15730853_transferring-information-or-data-to-a-cloud-network-server.html
http://www.123rf.com/photo_4916341_one-outstanding-question-mark-army-like-question-mark-formation-high-resolution-3d-render.html
http://www.123rf.com/photo_10365125_mayan-calendar.html


Time to Fly - Why Predictive Analytics is Going Mainstream

More Related Content

What's hot (20)

Similar to Time to Fly - Why Predictive Analytics is Going Mainstream (20)

More from Inside Analysis (20)

Recently uploaded (20)

Time to Fly - Why Predictive Analytics is Going Mainstream