Post on 13-Apr-2017
Copyright © 2015 Splunk Inc.
Jon Thies -‐ CTO, Qwasi ScoB Hall -‐ Sr. Technical Manager, NCR
Where Mobile meets In-‐Store & Point of Sale: Data Collides
Disclaimer
2
During the course of this presentaPon, we may make forward looking statements regarding future events or the expected performance of the company. We cauPon you that such statements reflect our current expectaPons and esPmates based on factors currently known to us and that actual events or results could differ materially. For important factors that may cause actual results to differ from those contained in our forward-‐looking statements, please review our filings with the SEC. The forward-‐looking statements made in the this presentaPon are being made as of the Pme and date of its live presentaPon. If reviewed aXer its live presentaPon, this presentaPon may not contain current or
accurate informaPon. We do not assume any obligaPon to update any forward looking statements we may make.
In addiPon, any informaPon about our roadmap outlines our general product direcPon and is subject to change at any Pme without noPce. It is for informaPonal purposes only and shall not, be incorporated into any contract or other commitment. Splunk undertakes no obligaPon either to develop the features
or funcPonality described or to include any such feature or funcPonality in a future release.
Agenda
! Speaker / Company Bios ! Why and How we uPlize Splunk ! SoluPon DescripPon ! Technical ExecuPon with Splunk ! Demo ! Q&A
3
IntroducPons
Scott Hall Sr. Technical Solution Manager - NCR
5
Jon Thies CTO - QWASI
6
?“CAN I TAKE YOUR ORDER?”
• A US-based software-as-a-service (SaaS) platform technology company.
• QWASI aggregates all real time customer
events so brands can create contextual experiences over Push, SMS, Email, Social, and IVR from one platform.
• QWASI’s scales and integrates with any
Mobile App, CRM, POS, Beacon, NFC, to activate all customer events for real time communication.
Real Time Data
Attribution, Analysis & Analytics
Behavioral Data
A unique Engagement and Data Management Platform that leverages real time data to curate unique user experiences.
Smart Engagement
Communication Protocols
QWASI’s consolidates first party behavioral data with CRM and Campaign Automation tools to create visibility into each individual’s journey. QWASI enables brands, agencies, and
enterprises to determine Campaign Attribution, ROI, and manage predictive protocols across all touchpoints.
How do we reach our customers?
9
! Do I start with a CRM? ! Geo-‐targePng ! Data filtering ! Real-‐Pme
Why Splunk?
! Splunk simplifies boring tasks – Monthly invoicing -‐ It would take an enPre day to review all the accounts – TroubleshooPng Customer Issues
! Splunk increases our ability to detect system failures and errors before our customers do
! Splunk saved our bacon – Our pladorm was new, our biggest customer was ready to leave for a
compePtor. The biggest feature gap in our pladorm was reporPng. – We created a Splunk App in two weeks that addressed every imaginable report
they needed.
Value of Splunk
11
Event Lifecycle
ApplicaPon Events
12
NoPficaPon Events Custom Events
ApplicaPons w/ Qwasi SDK
Point of Sale NCR Cloud Connect
Qwasi Cloud
Every InteracPon is an Event ! Applica'on Events – automaPc events generated using our SDK or API
– ApplicaPon Install/Open/Close – User RegistraPon – LocaPon – LocaPon Update, Geofence and Beacon Enter/Dwell/Exit
! Custom Events – events defined by Developers – For reporPng and analyPcs purposes – For triggering a workflow
! No'fica'on Events – events from QWASI to ApplicaPon via SDK or SMS – Sent – Delivery Report : Accepted/Delivered/Read – Failed
13
Why Splunk?
14
• Hundreds of clients • Thousands of users • Configuring over a hundred thousand campaigns • InteracPng with millions of subscribers • GeneraPng over a hundred million events monthly
Because its all about Events
Distributed event based pladorm -‐ combinaPon of dedicated servers, cloud instances, and Docker containers. ! Gain insight of usage at broad and granular levels
– Passive monitoring across data centers and deployments – System Health and Usage Monitoring
! Historical ReporPng and AnalyPcs that will not impact runPme systems ! Embed a flexible Splunk Dashboard into our user interface using SplunkJS
Why Splunk?
15
Because its all about events
Every Event Needs Context ! Events are always 'ed to a device and member profile
! NCR’s Cloud Connect – Point of Sale Integra'on – Customer Loyalty Data (profile) – Eligible RedempPons and Bonus Offers – Purchase History
! Event Triggers – workflows that fire as a result of a specific event within a predefined context.
! Scheduled Triggers – workflows that review the state of members profile and event history and iniPate workflows based on a schedule
16
SoluPon DescripPon
How the SoluPon UPlizes Splunk
18
Splunk Indexes:
• All Transport Traffic ê GCM, APNs ê SMS SMPP/HTTP ê MMS MM7 ê …
• All API Traffic ê SDK ê HTTP ê Loyalty APIs
• Each Event ê Member ID ê Customer ID ê Deployment ID ê Event Data Fields
How the SoluPon UPlizes Splunk
19
• Use of Scheduled Saved Searches • IT and OperaPons Team login to Splunk Web for troubleshooPng and support items daily.
• Provide Rich AnalyPcs and Insight into Campaign Performance and Customer Engagement
Ø BROWN Hot963: You're in to win tickets to see Chris Brown. Stay by the phone Reply STOP BROWN to Cancel; HELP=Help; STOP=End; Msg&Data rates may apply
DemonstraPon
Technical ExecuPon
The Data
22
• Events Log : /vol/aim_events.log As each event occurs in the pladorm its wriBen out to a log file as JSON.
• Members Data – contained in MongoDB for each deployment.
Member Lookup Table
23
• AutomaPc Lookup on source • Saved Search
– Extracts Hosts – Saves to CSV – Executes Shell Script
• Shell Script – Loads CSV of Hosts – Calls Web Service on each Host – Populates Lookup Table
Member Lookup Table
24
• Adds EnPre Member Profile to each event as fields
• Saves on license volume • Lookup tables typically have
pre-‐defined fields – we worked around that
• Lookup table becomes indexed
Dynamic Reports Using SplunkJS
25
• Simple Dashboard containing : – TimeRangeView – DropDownView – SearchManager – ChartView – SplunkMapView – TableView – TextInputView – CheckboxGroupView
• SearchManager – InteresPng Fields (SelecPon 1&2) – Main Search – populates ChartView – Map Search – populates MapView – D3 Search – populates custom views
Screenshot here
Dynamic Reports Using SplunkJS
26
• Customize the Report – Labels – OperaPons
ê Count ê Sum ê DisPnct Count ê Average ê Standard DeviaPon
– Graphing OpPons ê Stacked for Bar and Area ê Heat Map for Table
Screenshot here
Dynamic Reports Using SplunkJS
27
• VisualizaPon to Choose From – Bar – Sankey – Line – Area – Pie – Column – ScaBer – Table – Map
Dynamic Reports Using SplunkJS
28
• Filters – Used for Drill Down – Customizes Splunk Search used
ConsideraPons Using SplunkJS
29
• Users ONLY get their events ê Limits source, customer_id within Splunk at search Pme
ê Limits concurrent searches ê Created when the User is Created via Splunk Rest API
• Drilldown Behavior ê Modifies field selecPon ê Generates new search string ê Difficult to manage ê Most power feature
• Downloading Events ê Is this something we want to do?
Future Enhancements
30
• Add more D3 or HighCharts VisualizaPons
• Download Search Results • Enabled Real Time VisualizaPons
ê This worked, but it leaked searches that would conPnue to run.
• Link from other Qwasi reports to the SplunkJS and prepopulate the Filters.
Summary
Summary
32
Campaign Automa'on
Loyalty Engagement
Point of Sale Transac'ons Business
Intelligence
Cross Channel Communica'ons
Data Management
QuesPons?
THANK YOU jon.thies@qwasi.com scoB.hall@ncr.com