Introduction and Challenge Overview

The Story Behind The Solution

This solution pattern extends Globex, a fictional retail store that has undergone modernization and already embraces Event-driven Ansible, Automation Controller, and Apache Kafka as a streaming platform for customer feedback on their eCommerce website.

Here’s an overview of the key components:

  1. Customer Sentiments Analysis:

    • Customer sentiments are analyzed through an intelligent application.

  2. Event Trigger with Event-driven Ansible and Automation Controller:

    • An event is triggered using Event-driven Ansible and Automation Controller.

    • Filtered messages are dispatched to respective category Product Managers for review on an in-house Chat System.

  3. Feedback System and Grafana Dashboard:

    • Product reviews from the feedback system are visualized on a Grafana dashboard.

    • Data-driven intelligent decisions are made based on this feedback.

  4. ChatOps System Integration:

    • The ChatOps system is utilized to enact actions like blocking a product or offering discounts.

    • An event is triggered using both Event-driven Ansible and Ansible Controller to implement the requested logic or operation through the chat system.

This integrated approach enables Globex to harness cutting-edge technologies for prompt feedback analysis, informed decision-making, and seamless interaction with ChatOps systems, fostering efficient product management and enhancing customer engagement.

The Solution

This solution is built on an Event Driven Ansible architecture leveraging Apache Kafka and ChatOps as powerful data streaming platforms to initiate events based on filters or chat commands. The process for handling customer product reviews is as follows:

  1. Customer Review Processing:

    • Customer product reviews are submitted and pushed to a Kafka topic.

    • The data undergoes moderation for foul/abusive language through a Review Moderation Service and is further analyzed for sentiment using a Sentiment Analysis Service (in this case, focusing solely on sentiment analysis).

  2. Serverless Eventing Architecture:

    • OpenShift Serverless Eventing architecture, utilizing brokers, sources, and triggers, facilitates the smooth flow of data across different systems, ensuring scalability and full decoupling.

  3. Data Persistence:

    • Moderated and suitable reviews are stored in the Globex Product Review Database (PostgreSQL).

    • Sentiment analysis scores are persisted in a Time Series Database (InfluxDB) for creating a Grafana Dashboard to track category performance over time.

  4. Event-driven Ansible Playbooks:

    • Event Driven Ansible, triggered by sentiments and feedback filters, executes playbooks for specific actions.

    • Messages are then sent to the respective Product Manager’s chat room through ChatOps (Rocket Chat) System.

  5. Real-time Decision-Making:

    • Product Managers can make real-time decisions by combining ChatOps messages with the Grafana real-time dashboard, using insights from the InfluxDB time series database.

    • Decisions may include stopping a specific product, offering discounts, etc.

  6. Command Execution through ChatOps:

    • Product Managers issue commands on the ChatOps system (Rocket Chat) for specific categories.

    • This triggers an event captured by Event Driven Ansible, which runs the respective playbook to execute the command (disabling the product in the database or updating the discount).

In essence, this integrated approach empowers efficient and data-driven decision-making for product management, leveraging the capabilities of various technologies seamlessly.