Querona Data Virtualization platform for Data Science teams as a self-service data preparation tool.
Every Data Science team needs to feed AI tools with data, easily prepare required data and publish the results.
The presentation was shown at The Heart of Warsaw, October 20th, 2017 on a AI & Data Science demo day.
2. In a data-driven organization, everybody who needs data
for decision making has freedom in accessing required data
Internal /
external data
A data-driven organization
3. 3
2
1 Identify data visionaries
Make all your data accessible from a single place by everybody
Empower all users
4 Invest in self-service data tools
A data-driven culture step by step
5 Hold employees accountable
4. Analytical maturity
Analytical maturity levels
Businessvalue
Descriptive
analytics
Diagnostics
analytics
Predictive
analytics
Prescriptive
analytics
What has
happened?
Why did it happen?
What will happen?
What can we do to
make it happen?
Traditional reports
Modern Data Science & AI
5. Predictive
analytics
Prescriptive
analyticsWhat will happen?
What can we do to
make it happen?
AI, ML and Data Scientists need data
All data for AI, Machine Learning and
Data Science must be easy
to find, combine and publish
Data needs in Data Science:
• training data for AI / Machine Learning
• real-time data for working AI models
• access to internal and external data
• simple and secure data mixing
6. Actionable analytics process
Hypothesis
Find required
data
Verify the
hyphotesis
on data
Introduce
a data-driven
business
process
Empowerment of business users in data management
leads to automated, fact based decision making
data
7. Gartner, August 2017, Market Guide for Data Virtualization
ALL DATA AVAILABLE IN
ONE PLACE
VIRTUAL ACCESS TO
DATA
SELF SERVICE FOR
DATA MANAGEMENT
Through 2020, 50% of enterprises will implement some form of data
virtualization as one enterprise production option for data integration.
Querona – Data Virtualization Platform
8. Your existing data sources
Virtual Database with
self-service data management
User-friendly front-ends
all data accessible in one place
power users can
connect, mix and publish
data views
unlimited access to data
for any use
Querona Data Virtualization Platform
Complete Logical Data Warehouse: ETL-free, self-service, Big Data ready, utilizing Apache Spark
9. Querona features
• All data available in one place
• Combine data from multiple
sources
• ~100 types of data sources
• Multi level data security
• Real-time data access
• Data available for any client or
AI tool
• Data categorization for
GDPR/RODO compliance
10. Why Querona
Data Virtualization (DV) is not a new idea but since 2016 Gartner has
considered DV as a key trend in Data Warehousing and Data Analytics
• Self-service → more people can use data
• SQL Server wire compatibility → compatible with any client tool
• Apache Spark bundled → „Big Data Ready” in 5 minutes
• Competitive licensing model → DV available for all companies
11. Use case: Data Science team empowerment
• Connect internal & data
• Self-service data preparation (join multiple
data sources)
• Cache slow and remote data
• Load data to Hadoop or the cloud
• Expose data for AI
• Verify hypothesis on data
Querona solves the primary need of Data Science teams:
all data present in one place
12. Use case: event based reaction
Event filtering
A big credit card
transaction
10.000 events daily → 100 actions
CRM tasks
Marketing list
Is a VIP
customer?
Credit card limit in
range?
Good
repayment
history?
Has received
an offer
recently?
13. Querona value for: event based reaction
internal data
Events
External data
Querona usage:
• Listen to events
• Add information from other sources
• Integrate data with AI
• Change rules at any time (self-service)
• Export results as CRM actions or
custom marketing campaigns
Querona closes the whole event lifecycle, from acquisition, through
augmentation and integration to publishing
14. Use case: data publication
Create custom “data views” for business partners or other departments
using the self-service data portal
Data publication for external use:
• AI recommendations
• Orders
• Billings
• Segmented customer profiles
• Complaints
15. Use case: automated recommendation
Orders
Products &
categories
Product recommendations:
• Frequently bought together
• Increase product discoverability
• Personalized recommendations
Train
Lookup a
recommendation
Querona integrates with Microsoft Azure Recommendations API
16. Customer profile in CRM
Full customer profile,
segmented customer profile
Secondary / legacy CRM
External information
(public, PSD2, …)
Customer behavior (sales, contacts
with the call center, etc.)
Digital marketing history
Use case: customer 360° view
17. Use case: GDPR/RODO compliance
Querona simplifies GDPR/RODO audit and enables
to maintain a long term compliance
GDPR/RODO compliance needs:
• Connect all data sources
• Look deeply into data from any source
• Identify personal and sensitive data
• Expose pseudo-anonymized data for analysis
• Export all customer’s data on request
18. Piotr Czarnas
CEO & Founder
Software Engineer
Experienced
Manager
Marek Byszewski
VP of R&D & Founder
Engineer
Experienced Enterpreneur
& Chief Security Officer
Robert Rogaczewski Irek Kalicy
VP of Sales
Executive MBA
Experienced Head
of Sales (Telco/IT)
VP of Operations
Engineer, MBA
Experienced CIO
Management team