Data analytics to improve home broadband cx & network insight
0 likes109 views
Collection of my ideas on #broadband Services #cxtransformation. I feel #dataandanalytics will definitely help in close loop systems, not only in #networks but also the business processes. #customerexperience is the key for ISPs & CSPs for retention and loyal customer base. The current network are improving the data set availability by using #telemetry #USP and #netconf, but still lot more standardisation is needed in this area, iOAM can be great protocol to implement. The Linux Foundation is also there in data analysis and AI, really thankful to them for democratisation of technology. #PNDA #ACUMOS #aiforeveryone #dataanalytics #closedloop #broadbandnetworks #ftth #NLP #predictiveanalytics #prescriptiveanalytics #analytics #analyticsplatform
Data analytics to improve home broadband cx & network insight
- 1. Data Analytics to improve Home broadband CX & Network Insight Ravi Sharma
- 2. CUSTOMER EXPERIANCE Customer experience(CX) is the product of an interaction between an organization and acustomerover the duration of theirrelationship. This interaction is made up of three parts: the customer journey, the brand touchpoints the customer interacts with, and the environments the customer experiences (including digital environment) during their experience. A good customer experience means that the individual's experience during all points of contact matches the individual's expectations. IMPROVE - Process, People & Technology CUSTOMER JOURNEY DIGITAL ENVIRONMENT HOW TO IMPROVE INTERACTION TOUCHPOINTS MEASURE RESPONCE RESOLVE Customer DATA FEEDBACK
- 3. A P P/ W E B / C A L L C E N T R E Digital Environment O 22 Data Analysis Data Set CX Implementation - Home broadband Use Cases 1 2 3 Identify Use Case Check data available to implement use cases Start the Data analytics Use Cases shall cover these CX parameters CUSTOMER JOURNEY INTERACTION TOUCHPOINTS O R D E R / S A L E S I N S TA L L AT I O N U S A G E /A P P S B I L L I N G 1 2 3 4 A P P/ W E B / C A L L C E N T R E H O M E V I S I T C P E / S T B / S E RV I C E A P P S A P P/ W E B / C A L L C E N T R E A P P W E B C A L L C E N T R E S O C I A L M E D I A S EA R C H E N G I N E S C O M P L A I N T S5 N E T WO R K 1) Networks are complex to get the data related with customers 2) Usages , Billing & Complaints data set is available but correlation of these data sets are required
- 4. M More on Analytics Platform Data Data Source Visualisation AI ML& Business Insights & Value Deep Learning Data Pipeline Data Ingestion Analytics Platform Multiple tools depends On parameters As Frequency, size Multiple Types Depends On Source & type of data I.e. Message Brokers APPS WEB Social Media Network CPE CRM Ticketing Data Collection: Data Storage Data Access AI ML module Deep learing Module Curation & Search I N T E G R A T I O N D E V O P S S E C O P S Visualisation Tools AI ML Tools
- 5. M More on the data analysis Data Visualisation AI ML& Business Insights & Value Deep Learning REPORTING VISUALISATION DESCRIPTIVE ANALYSIS DIAGNOSTIC ANALYSIS STATISTICS MODEL PREDICTIVE ANALYSIS MACHINE LEARNING MODEL PRESCRIPTIVE ANALYSIS MACHINE LEARNING MODEL DATA INPUT REPORTING INSIGHT PROACTIVE COGNITIVE The e鍖ective combination to automate the feedback and Self learning GCP ML
- 6. M More on the data analysis Arti鍖cial Intelligence Machine Learning Deep Learning Source: Image authored bySherwin Chen Machine Learning is the subfield of computer science that gives computers the ability to learn without being explicitly programmed. ~ Arthur Samuel Deep learning is a subset of machine learning in which layered neural networks, combined with high computing power and large datasets, can create powerful machine learning models. Image source:http:// www.cognub.co m/index.php/ cognitive- platform/ U S E C A S E S A L G O R I T H M S
- 7. Linux Foundation platforms Acumos AI is a platform and open source framework that makes it easy to build, share, and deploy AI apps. Acumos standardizes the infrastructure stack and components required to run an out-of-the- box general AI environment.
- 8. M How to improve- Customer Journeys & Interaction CUSTOMER JOURNEY O R D E R / S A L E S I N S TA L L AT I O N U S A G E /A P P S B I L L I N G 1 2 3 4 C O M P L A I N T S5 FULLY DIGITAL CUSTOMER JOURNEYS PROVIDES DATA POINTS MEASURE RESPONCE RESOLVE Customer DATA FEEDBACK INTERACTION TOUCHPOINTS A P P/ W E B / C A L L C E N T R E H O M E V I S I T NLPAI Location Awareness Integration of Chat bot & Map services in Operators mobile app E-Commerce Website like exp 1- Order 2- Delivery 3- Complains
- 9. M How to use data points - Orders/Sales process Ful鍖lment timeOrder MEASURE The data points Just an EXAMPLE Install time Lead time Payment success App experience Journey Data Analysis Outcome Feedback to improve The process Close loop systems are required to improve the process & touchpoint Products Insight Products launch
- 10. Trends Network Analytics Lets start with use cases again Network Topology Physical Logical Geological Traf鍖c Routing Provisioning Inventory Physical Active Redundancy Inventory logical Customer Complaints trends & auto resolution Service Quality prediction Internet Peering Bandwidth Utilisation trends Capacity Planning Data usage Bandwidth, Delay etc Operations Planning Service NOC Network NOC Automated provisioning Fault prediction Anomaly detection Alarm trends Business Revenue leak Churn Homes Wi鍖 Performance trend Channel Change List goes on
- 11. Are Network Analytics Get the data ONT AAA DHCP DNS Core BNG OLT IP NW PCRF TR 069 SNMP OMCI SNMP NETCONF IOAM TELEMETRY SNMP NETCONF IOAM SNMP NETCONF BGP Network elements can provide the data to a centralised data analytics platform such as PNDA. Operators are also building there own data warehouse/ data lake based on open source technologies. The data analytics platform provides the input the DATA VISUALISATION tools such as Tableue or Operators build custom GUIs. Visualisation Tools/ AI ML / Close loop
- 12. Are Network Analytics Source: CiscoDevNet/iOAM Di鍖erent close loop in an automated environment
- 13. Are Network Analytics Di鍖erent close loop in an automated environment Source Intel Corporation.
- 14. Are Network Analytics Few use case in details Optical Health Collecting optical parameters (seconds) Pattern recognition based on SOP Supervised learning based on earlier faults Change the path Before interruption Stanford Paper : Link Failure prediction and Localization in Cloud Scale Networks using Supervised Learning Here is the link to the zip 鍖le for this project: https://github.com/mehrnaz22/ML_project/blob/master/ML_Project.zip
- 15. Are Network Analytics Few use case in details Anomaly detection Supervised (Semi) learning Collecting optical parameters (seconds) Pattern recognition based on SOP Change the path Before interruption > Learn the time series pattern from historical data (and labels) > Predict future values and compare with real measurements Anomaly Detection is a ML process to distinguish signal from noise, the same can be applied to alarm received from the network The method to implement is semi-supervised learning Firstly the model is trained on actual data is like a supervised regression model. In the second stage, Anomalies predicted whenever actual value breaches either upper or lower bounds is a representation of unsupervised learning model. though the second stage is unsupervised, the user can manually reset the values to adjust to the signal accordingly. https://web.eecs.umich.edu/~zmao/eecs589/papers/AnomalyDetection.pdf https://blog.cloudera.com/deep-learning-for-anomaly-detection/
- 16. One cannot think well, love well, sleep well, if one has not dined well. Virginia Woolf