Module 7 Flashcards
Enterprise systems and data driven decision making
What is TPS?
Transaction Processing Systems: record data on fundamental operations occuring within the company.
Types of TPS
- Batch processing
- Online transaction processing (OLTP)
Batch processing
- Stores data in a temporary location and is processed in single units (batch) at a specific time
- Allows companies to shift processing to times when computing is less busy
Online Transaction Processing
- Data is processed in real time
- Current state of the system is always reflected
How do enterprise systems provide value?
They provide value by increasing operational efficiency and by providing firmwide information so managers can make better decisions.
Enterprise software
Enterprise software includes analaytical tools to use the data the system captures to evaluate the overall organization performance.
What is ERP?
Enterprise Resource Planning (ERP) intagrates functions of the company into a homogenuous system. Leaders are SAP and Oracle.
What is a supply chain?
A supply chain is a network of organizations and business processes that provide the company with raw materials, transform these materials into intermediate and finished products and deliver the finished products to the customer.
Downstream portion of supply chain
Includes organizations and processes for distributing and delivering end product to final customer.
Upstream portion of supply chain
Includes suppliers and processes for managing relationships with them.
JIT
Just-in-time strategy is an inventory management method, in which materials are received by suppliers only as they are needed, in order to lower inventory holding costs and increase inventory turnover.. This can be achieved when we have perfect information, which is not possible, so that’s why companies hold Safety Stock, which acts as a buffer for the lack of efficiency.
Bullwhip effect
This happens when information about the demand for a product gets distorted as it passes through the supply chain. Small change in demand gets magnified, which leads to exess inventory.
Supply chain planning system
These systems model existing supply chain, make demand forecasts and develop a plan for optimal sourcing and manufacturing.
Demand planning
Determines how much of a product a company should produce to satisfy all customer demands.
Supply chain execution system:
Manage the flow of products through distribution centers and warehouses to ensure the product gets delived to the right location in the most efficient manner.
Push-based model
Production master schedule is based on forecasts of customer demands and product has to be pushed to customer.
Pull-based model
Actual customer purchases trigger events in the supply chain.
CRM systems
Customer relationship management systems integrate data from various departments.
CRM packages include tools for:
- Sales force automation
- Customer support
- Marketing
Touch point
Method of interaction with the customer: email, FB, CS desk
Sales force automation
SFA means making sales staff more productive by focusing sales efforts on most profitable customers.
Operational aspect of CRM
Customer facing application
Anlytical aspect of CRM
Applications that analyze customer data to provide information for improving business performance.
Output: Customer lifetime value CLTV
Management Information Systems
- serve middle managers
- answer routine questions with predefined procedures to get to an answer
- little analytical capability
- provide reports on firm performance based on TPS
Decision Support System
- serve middle managers
- can use external resources as well as MIS/TPS
- support non-routine decision making
Executive Support System
- support senior management
- non-routine decision makinh, requires judgement, evaluation and insight
Data warehouses
Collect and store data from transactional systems across the organization. Data is consolidated and standardized, but can’t be altered. These serve as a tool for querying, analysis and reporting.
Data marts
Are subsets of data stored in data warehouses:
- contain a highly focused proportion of the organization’s data
- designed to support a specific business problem/challenge on a specific population of users.
Data lakes
Centralized repository that allows you to store structured and unstructured data at any scale.
Contemporary business intelligence infrastructure
Combines “traditional” data warehouse solutions with data lakes to integrate unstructured data and/or external data and/or large structured data sets.
What do database, data warehouse, data marts use as input?
Operational systems like TPS.
What is the output of databases?
Business intelligence tools, analytical systems.
What are business intelligence tools?
Tools for consolidating, analysing, accessing data to support organizational decison-making.
Online analytical processing
OLAP: transaction-level data from relational databases is aggregated and summarized.
Where are the results of analysis from OLAP stored?
They are stored in special databases called data cubes, which structure results across multiple dimensions. (Time, Place, Product)
Data Mining
The use of special algorithms to identify hidden patterns in and to fit models to large data sets.
Association
Certain atrribute values that frequently occur together in a data set. (enables vendors to tailor discounts and optimize product location)
Central concpets on association
- Support x(X) is the fraction of transactions that contain a certain set of items X.
- Confidence c(X->Y) is the fraction of transactions that contain Y amongst the transactions that contain X.
Big Data
- Volume: The simple size of a dataset that needs to be processed;
- Velocity: The speed with which new data is generated and needs to be processed;
- Variety: The different formats & features of data that need to be processed (relational databases, documents, photos, videos, spatial & temporal aspects);
- Veracity: The reliability of the data (user-generated content, such as restaurant reviews, but also imprecisions in measurements, such as GPS positions).
Analytics
Analytics applies classical statistics and AI to derive actionable insight from big data. e.g: neural networks replicate the basic functions of the human brain to predict outcome
Neural networks
- Are feed historic data sets on the outcome of interest or other variables
What did Siemens implement and how>
Implemented Celonis’ Business Mining software.
- analyses data in enterprise application event log to predict bottlenecks and other inefficiencies
e.g: using special algorithms it can predict which orders will be late based on historical outcomes of Siemens
- previously individual supervisors were responsible for one process making it difficult to understand which processes were affected by one process
- organisational resistance to process mining from long-term managers
Types of decisions
- unstructured decisions: decision-maker must provide judgement, evaluation and insight on the decision. no well-understood procedure on makin them
- strctured: repetitive and routine, definite procedure
- semi-structured: only part of the problem has a clear-cut answer provided by an acceppted procedure
