- A Process Improvement technique: Plan Do Check Act

Filter Navigator - System Diagnostics - An entire application in ServiceNow to look for root causes of performance issues.
- Stats ServiceNow Performance Home Page

ServiceNow Performance Home Page 
- Why would the chart only have data since July 11th? \\

- Transition Log by Response Time and other values. Sort “Descending” to see the worst offenders. Transaction Log
- Diagnostics Page System Diagnostics page
- Articles https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB0517241 Troubleshooting Slow Performance
https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB0517282 Troubleshooting General Performance
Issues on all Applications https://support.servicenow.com/kb?id=kb_article_view&sysparm_article=KB0517269
Gathering Stats.do and Threads.do Page Data https://docs.servicenow.com/bundle/quebec-platform-administration/page/administer/core-configuration/concept/p_PlatformPerformance.html?cshalt=yes
Platform Performance page. Very useful.

- Takeaway Performance Analytics (PA) is a system for analyzing data to help with process visibility and improvement. PA takes a “snapshot” of data each and every day (using a scheduled job). This gives a better understanding of the history of each record. Also, unlike regular reports, PA involves the establishments of Targets and Thresholds and can help Forecast future performance. The line between reporting and analytics is not always clear, but generally speaking, reporting shows current state and analytics shows trends. When implementing Performance Analytics you define Indicators/KPIs/Metrics (what to measure, and how often to measure) and Breakdowns (grouping variables), then you collect data in Scores tables, then visualize the Scores and Trends on a Dashboard. “Analytics Solutions”, also known as “Content Packs” are pre-packaged bundles of best practice KPIs, Widgets, Dashboards, and other supporting content to support various Processes.
- Modules

Performance Analytics - Analytics Hub
Performance Analytics - Dashboards
Performance Analytics - Data Collector - Jobs - New
Performance Analytics - Data Collector - Job Logs
Performance Analytics - Sources - Indicator Sources
Performance Analytics - Indicators - Automated Indicators
Performance Analytics - Indicators - Thresholds
Performance Analytics - Breakdowns - Create New
Reports - Administration - Report Sources  \\

- Plugins

com.snc.pa.premium - Needed for use with Custom Applications \\

- Roles

dashboard_admin - Can edit and manage users, groups and roles for Any dashboard.
pa_admin / pa_power_user - Can manage users, groups, and roles on dashboards to which they have Edit permission.
pa_data_collector - Can create Indicator Sources.
pa_target_admin - Can create Global Targets.
pa_threshold_admin - Can create Global Thresholds.
pa_viewer - Can create, edit, and delete their own dashboards.  Can create Personal Targets and Thresholds. \\

- Performance Analytics is a tool for analyzing data over time, rather than reporting against only a point in time. - Performance Analytics is an information presentation tool which helps identify bottlenecks, and improve the efficiency of your processes and services. - PA is for Business Process Performance Management. It delivers real-time insight into business performance. It helps reduce the cost of service delivery. - Theoretically you can optimize staffing levels every day so you can respond faster to requests, across all lines of business. - Performance Analytics ships with pre-configured dashboards and KPIs tailored to support a variety of processes. - Operational Reporting often looks only at the current state of a record, and does not fully grasp the history of each individual record, such as how long it sat in one assignment group before being transferred to the current assignment group. - PA takes a “snapshot” of data each and every day (using a scheduled job). This gives a better understanding of the history of each record. - PA helps track performance against Targets, alerts when Thresholds are met, helps Forecast future performance, and helps compare performance at different points in time. - The line between reporting and analytics is not always clear. Generally speaking, reporting shows current state and analytics shows trends. - PA System Tables are known as Scores Tables. Analytics are often used to predict future outcomes or to improve processes. - All ServiceNow instances are provisioned with an unlicensed version of Performance Analytics. This version has configuration limitations, such as a limitation to not preserve Scores for longer than 180 days. - A license is required to enable the complete set of Performance Analytics features. - When implementing Performance Analytics define indicators (what to measure) and breakdowns (more in depth analysis), then collect scores, then visualize the scores and trends. - PA should be built into an App starting from the data model design. Avoid string fields. Be sure fields needed for analytics are mandatory. - Common pitfalls (without PA): Measuring everything. Displaying values without the context of Time or operational targets. Metrics hidden in PowerPoint decks. Dashboards overloaded with content but failing to provide useful decision-making guidance.
- Analytics Hub

1. The User Interface for assessing, comparing, and predicting the performance of indicators over time.
2. Displays data for a single indicator at a time.

- An exploratory view of an Indicator, used for more detailed analysis. It lets you see trends, predictions, breakdowns, and the associated records for one specific Indicator. Involves statistics, forecasting, and comparison. - To see indicators in the Analytics Hub, the indicator must be Published. - Use the Compare tab to compare the scores of an Indicator from two measurement periods. - You can manually add comments to a chart to call out important features. - Analytics Hub replaced “Scorecards” starting with the Madrid release of ServiceNow.
- Breakdown

1. A more in-depth analysis of an Indicator.

- A grouping of a particular Indicator, with respect to another variable. For example, Open Incidents may be broken down by Priority; or by Assignment Group. Allows you to view data by a specific attribute.
- Content Packs: - Out of the Box Analytics solutions are provided as Content Pack plugins. - “Analytics Solutions”, also known as “Content Packs” are pre-packaged bundles of best practice KPIs, Widgets, Dashboards, and other supporting content. They are available for many ServiceNow Applications to help you get started quickly. A list is available here: Out of the Box PA Solutions - Deployment Steps:

1) Activate Analytics Solution / Content Pack
2) Validate Indicator and Breakdown Sources
3) Configure and Run Collection
4) Configure Dashboard Permissions and View Dashboards \\

- Data Architecture / Data Layer

	1)  Indicator Source
		- Example:
			Name:				NeedIt.ThisMonth
			Table OR Report Source:	NeedIt OR Active NeedIt Requests
			Valid For Frequency:		Monthly
			Conditions:			When Needed ON This Month

- Defines a Filtered set of records, collected at a regular Frequency. - Specifies a set of records from a table with a common characteristic, such as Priority = Critical.

1. Frequency

- Include a date-related filter also, such as “Opened on Today” or “Closed on This month”.

1. Not easily changed, since these are used by multiple Indicators.
2. Has a Report Source or a Facts Table and Conditions.
   1. A Report Source is a filtered set of table data.

- Performance Analytics - Sources - Indicator Sources

	2)  Indicator / Key Performance Indicator (KPIs) / Metric
		- Example:
			Name:				NeedIt Due Monthly
			Frequency:			Monthly
			Direction:			None
			Unit:				#	(count)
			Indicator Source:		NeedIt.ThisMonth
		- Defines "What to Measure" and "How Often" to measure (but not the set of records to measure from).

- Defines a specific measurement, you can count or calculate, to assess process performance. Example: Number of open incidents. - Types:

1. Automated - Scores are automatically collected using a scheduled data collection job on a regular frequency.
2. Manual - Scores are entered manually or imported from a third-party source.
3. Formula - Scores are calculated using scores of other indicators.

- Properties:

1. Direction - The preferred direction of the Indicator: Maximize or Minimize, or None (does not matter).
2. Collect Records - When selected, stores sys_ids for individual records, enabling later drill down.
   

3) Data Collector

1. Example:

Name: NeedIt Daily

			Relative Start:			1
			Relative Start Interval:		Days Ago
			Collect:				Scores Only
			Time:				1:00 AM
			Indicators:			NeedIt: Due Monthly

- Scheduled Job that collects data from an Indicator Source (specified indirectly by an Indicator) - Quirk: To run a Data Collector on the last day of every month, create a monthly execution set to run on day 31. It will automatically run on the last day of the month, even for months with fewer than 31 days. - Tip: Create an Admin user with the “Web service access only” setting selected. This setting does not allow the user to log into the ServiceNow user interface. Use this user to run the Data Collector. (Not sure why this is a good idea.) - An Indicator may be added to a Data Collector at any time. The data collector collects new indicators beginning with the next collection period. - A Historical Data Collector may be defined to collect data for existing (old) records. Run Historical Data collection when PA is first setup for an Application, Indicators Sources are created, or Indicators are created.

	4)  Breakdown Source
		- A set of records from a table or database view.

- Defines the list of breakdown elements a Breakdown contains. - Collects unique values from a filtered set of records. - Multiple Breakdowns can use the same Breakdown source. - Appears to consist of both an Indicator on a Table, and a particular Field. - Bucket Group

1. A special type of Breakdown Source that divides values returned by a script into discrete buckets.
2. Useful for drawing conclusions that may not be obvious in the raw data.
   

5) Breakdown

1. Example:

Name: NeedIt Request Type

			Indicator:	NeedIt Due Monthly
Field:		Request Type

- Filters or groups Indicator Scores for detailed analysis. - Can be used for multiple indicators, based on different facts tables. - A Country breakdown could be used for indicators from the incident, change, and earthquake fact tables.

1. Visualized in Analytics Hub and Dashboards.
   

- Dashboards

1. Dashboards organize PA widgets into logical groupings to provide insight into application data and performance.
2. A Drag and Drop canvas for gathering visualizations used regularly into a single location.

- Use both PA and Reporting widgets to build a comprehensive picture of your business process health. - You can create widgets from scratch in just a few minutes because there are existing indicators in the system that you can select as your starting point. - Tip: Create a dashboard for each Application Persona.
- Data Collector - The engine that takes periodic snapshots of your process tables and stores those snapshots in the Scores and Snapshots tables for later analysis. - Forecast

1. PA can forecast future data in a Time Series Widget.
2. A Forecast appears as a dotted line on a chart.
3. Forecast methods include: Linear, Drift, Naïve Seasonal, Naïve Seasonal Drift, Seasonal Trend Loess, and Random Forest.
4. The “Auto” method uses all other 6 forecasting methods, and then selects the one with the best fit.
   

- When viewing a chart in Analytics Hub, if the user clicks the Forecast button, but no Forecast is defined in the indicator record, Analytics Hub will apply an Auto Forecast for the indicator.
- License

1. The PA Premium license allows you to:

- Create indicators, breakdowns, and other records - Create interactive filters and use interactive analysis - Create text analytics widgets - Use Performance Analytics with external data - Preserve scores beyond 180 days
- Personas typically interested in PA:

1. End Users
2. Workers / Fulfillers
3. Service Owners

- Executives
- Score

1. I believe a Score is a particular sampling of an Indicator.
2. When you pull data from a table at a particular point in time, you record a Score.
   

- Targets

1. A Target is a goal that an organization wants to achieve using an Application.
2. Use a Target to visualize the delta between a desired Score and an indicator Score.
3. A Personal Target is visible only to the user who created it, and appears as a light line.
4. A Global Target is visible to all users and appears as a dark line.
5. When a Target is reached, PA generates a Notification.
   

- Thresholds

1. Define a normal range of Scores for an Indicator; may be Personal or Global.
2. Thresholds appear as dotted lines; light gray for Personal, and dark gray for Global.
3. Send a notification when certain events occur, such as when a Score reaches an all-time high or low.
   

- A job role I should consider targeting. - A Senior Leader with overall accountability for the Now Platform.

- Takeaway - A Plugin is a ServiceNow software component that may be optionally activated. Keeping unnecessary Plugins turned off reduces menu clutter and improves system performance. Some Plugins require a paid Subscription before activation. Many Plugins include Demo Data. Tables

sys_plugins	(Use this)
v_plugin

- Modules - System Definition - Plugins - A “Plugin” is a software component (disabled by default) that provides additional, optional, features and functions within a ServiceNow instance. - By NOT having plugins turned on, you reduce the complexity of the system. You will have fewer menu items, for example. And performance will be better. - Note however, you can hide enabled items by making them Not Active or by removing all Roles (except system administrator). - If a Plugin does not show up under System Definition - Plugins - All Applications, it may require activation by ServiceNow personnel. - Some Plugins require a paid Subscription before activation. Contact your ServiceNow account manager. - Demo Data - Many plugins include Demo Data; these are sample records that are designed to illustrate plugin features for common use cases. - Demo data should be loaded only in Development or Test instances. - It sounds like the v_plugin table only lists plugins which are “self-service”; I think that means those which you can activate yourself with no help from ServiceNow… And the sys_plugins table is the more reliable table to see what is actually installed and activated. (I did notice that both tables contain some unique values however, so I guess to be 100% comprehensive, you would need to check both tables. The “All Applications” module seems to be just another UI into the v_plugin table. However, unfortunately, the terminology used in not consistent…

              Installed == Active
              Not Installed == Inactive

- Activate Plugin - This is done from the ServiceNow Service Portal after you drill into a particular Instance.

-Banner Frame - A horizontal strip that appears at the top of every page and contains:

Left:	Logo

Right: User Menu (with user's picture), Global Search, Help, Connect Chat, and System Settings. Impersonate User and Logout are available here. In the Search Bar, Boolean Operators such as AND & OR must be entered in ALL CAPS. Search Results are organized by “Group” and “Table” and include only records the logged in user has access to, by Role. The Search Results “Summary” on the right shows the number of record matches under each Group and Table. Underneath Help (question mark icon) the hamburger menu provides access to the User Guide and Product Documentation. - System Settings (Gear Icon) is where you can personalize any part of the system for yourself, without affecting other users. - Application Navigator / Filter Navigator - A vertical strip on the left that provides access to all Applications (“Application Menus”) and nested “Modules” available to the logged in user. An Application is really just a “Section” or “Category” or “Sub-Menu”. - Module Filter - The top search control. - Hidden Feature: Typing TableName.List or TableName.Form into the Filter Navigator is a shortcut to display the table in list or form view. This can be done even if no Module exists for the table in question. - Clicking the Star on the right side of any Module adds it to your favorites list. “Type-Ahead” is the name of the feature where a filter is applied instantly as you type, without the need for pressing the enter key. - Tabs - All Applications - Hidden Feature: Double Click the All Applications tab to quickly expand or collapse all modules displayed. - Favorites - Hidden Feature: You can add a Filtered list to Favorites by dragging the Breadcrumb listing to Favorites. - Hidden Feature: You can add an Individual record from a list to Favorites by opening Favorites and then dragging the record Number to the Favorites left panel space. - The Edit icon (bottom right of Favorites) allows you to change the name of your favorites. You can change colors as well. - Simply Drag and Drop items to re-order them. - History

1. Shows the modules or system functions that you have recently accessed, and allows you to return there. Often overlooked feature.

- The navigator may be collapsed to the left, to free up screen space for the Content Frame. - Hidden Feature: You can hold down the Ctrl or Shift keys before clicking on a Module in the Filter Navigator. This will launch your selection in a new Tab or a new Window, respectively. - Content Frame - Main frame which displays context-sensitive content from Dashboards or Modules.

1. Displays Lists, Forms, and other types of content.

- Clicking on a Module displays its information in the Content frame. - List View. See “Lists” in this document for much more about lists displayed in the content frame.

- Takeaway - Predictive Intelligence (PI) is a subscription-based artificial intelligence solution in the Now platform. It learns from patterns in relatively large quantities of historical data records, becomming increasingly accurate in its predictive recommendations. Complicated to setup and requries at least 30,000 recoreds to work from. PI “Frameworks” allow predictions based on Classification (grouping), Similarity, and Clustering. “Coverage”, the % of records that a PI can make a prediction for, is inversely proportional to “Precision”, the % of predictions that are correct. A Solution Definition involves specifying a table containing the records you want a Prediction on, the Field you want to populate, the Input Fields you want the solution to consider, and a Filter condition to exclude irrelevant records. ServiceNow provides several Out-of-the-box Solutions for Incident, Knowledge, HR, Cases, and Event. Training is done by sending data to an external Training Server. The Training Server produces a Trained Model which may then be used (directly?) on the customer Instance. A Prediction Service may offer predictions as a service. - Personas:

System Administrator					- Can implement and manage all aspects of PI.
Application Administrator	- ml_admin 		- Manages the PI Application.
Report User			- ml_report_user 	- Can view dashboards.

- Plugins:

com.glide.platform_ml

- Modules:

Predictive Intelligence - Similarity - Solution Definitions
Predictive Intelligence - Classification (?) - Solution Definitions
Predictive Intelligence - Clustering - Solution Definitions
Predictive Intelligence - Word Corpus

System Web Services - REST - REST API Explorer - Predictive Intelligence is an artificial intelligence solution embedded within the Now platform. - AI is the general concept of machines acting in a way that mimics human intelligence. Could involve communication or decision-making. - Learns from patterns in historical data, becoming increasingly accurate in its predictive recommendations. - AI Algorithms need to be trained over time, with large volumes of well-organized data. - There are several properties that you can configure for Predictive Intelligence. To see a complete list of properties, navigate to the System Properties table and filter records where the Package field is Predictive Intelligence. - PI does require a Subscription. - ServiceNow provides a REST API Explorer utility. - Each data center contains the following types of servers:

1. Trainer Server - Generates a model based on the data that you defined in the solution definition.
2. Prediction Server - Offers predictions as a service.

- Note: There seems to be quite a lot that can go wrong with Predictive Intelligence, and you need large data sets. Also, there are a number of moving parts. I can see this being quite difficult to setup, to debug, and to maintain. - PI is not supported if you are using “Edge Encryption”. - API

1. PI has many different server-side scripts that you can use to customize its behavior.

- To view the complete list of scripts for PI, navigate to System Definition > Script Includes and apply the following filter condition: Package | is | Predictive Intelligence. - Commonly used script includes:

ML Predictor - To find and apply predictions on a table.
	- Get Predictions
	- Has Detailed Outcome
	- Detailed Outcome
ML Solution Result - To look up ML solution results.
	- Get Cluster Info
	- Get Cluster Assignments

REST API Methods:

Prediction (GET) 			- Predict an output field value using a specific solution.
CancelSolutionTraining (GET) 		- Cancel request for training.
GetRecordsForSolutionDefinition (GET) 	- Retrieve records meeting the filter conditions for a specific solution.
CreateTrainingRequest (POST) 		- Initiate training request.
Auto_train (POST) 			- Train default solutions.
GetOOBSolutionDefinitions (GET) 	- Get a list of solution definitions that are out of box.

- Business Case: - Helps optimize resources and processes, improve customer satisfaction, decrease the amount of time it takes to resolve issues, and ensures energy is focused on innovation and high-value activities. - Helps reduce assignment error rates. - Helps reduce the number of required interactions by correctly assigning manual tasks and automating tasks when possible. - Helps fulfillers quickly find similar open issues and determine if escalation is required. - Class List

 - A list generated by PI, of the possible values for a specific output field (based on a data set).
- There is no limit to the number of classes that a Classification model can support.

- Coverage

1. The percentage of records that an ML Solution is able to fully process, and make a prediction for.
2. When a Solution is trained, an estimate of Coverage becomes available, and is known as Estimated Coverage.
3. Increasing coverage will result in a lower Precision. Lowering coverage will result in higher Precision.

- Frameworks

1. ServiceNow Predictive Intelligence offers 3 different types of “Frameworks”. Each Framework has its own data tables.
2. Framework Types:
   1. Classification
      1. Deciding what group to place a record into.
      2. Use historical data to accurately categorize (and then route) tasks, incidents, and cases.
      3. An ML Solution that sets field values during record creation, based on user inputs.
      4. For example, the Incident category can be set based on the Short Description field.
   2. Similarity
      1. An ML Solution that identifies existing records that have similar values to a new record.
      2. Identify similar incidents, cases, and alerts to predict new major incidents and recommend actions.
   3. Cluster / Clustering
      1. An ML Solution that groups similar records into clusters so you can address them collectively or identify patterns.
      2. Supports English only.

- Learn

To progressively improve performance on a specific task, using data, and without being explicitly reprogrammed.

- Machine Learning (ML)

1. The ability of computers to learn from experience.
2. Machine learning algorithms are typically designed for specific tasks such as facial recognition or malware detection.
3. Algorithms can be “trained” on large volumes of data.

- Natural Language Understanding (NLU)

1. A branch of AI that uses computer software to understand input in the form of sentences (in text or speech).

- Precision

1. The percentage of predictions that are correct for a Machine Learning solution.
2. When a Solution is trained, an estimate of Precision becomes available, and is known as an Estimated Precision.

- Process for Implementing: - Tips for Success

	1) Identify the problems you want to solve.
	2) Determine if you have enough data from which PI can learn (30,000 to 300,000 historical records).  Minimum # is 10,000 records.
	3) Determine if your data is Correct.
	4) Understand that PI is imperfect, and there needs to be a manual backup process.
- Timeline:
	Day 1:
		- Clone production to sub-prod.
		- Install PI plugin on sub-prod.
	Day 2-10:
		- Create solution definition.
		- Train.
		- Validate / Tune Solution.
	Day 11-13:
		- Install PI Plugin on prod.
		- Create solution definition (from update sets)
		- Train
		- Validate / Tune Solution.
		- Set Training Frequency.
	Day 14:
		- Monitor Solution

- Solution Definition

1. You specify:
   1. A Table containing the records you want Prediction on.
   2. The field you want to populate.
   3. The Input fields you want the solution to consider.
   4. A filter condition, to exclude irrelevant records.
2. The end goal is to maximize the accuracy of the model against the most useful data.
3. A Solution Definition can have multiple Solutions, however only one Solution can be active at a time.
4. Your data set should reflect the current business process, and active and valid field values.
5. Before training, remove (filter out) data such as machine-generated records that you do not want the model using to make predictions.
6. Be sure you choose the right, and relevant input fields.
7. Retrain your PI solutions any time your business process changes. Do this manually, or on a scheduled basis.

- Solution / Trained Solution

1. A Trained Machine Learning Algorithm.
2. Contains the underlying Predictive Models.
3. Generated when a solution goes through the training process.
4. Can be invoked by any application using the “Prediction API”.

- Out-of-box solutions that ServiceNow Provides: Application / Process Solution Definition Framework Type Description IT Service Management (Incident Management) Incident Assignment Classification Predicts the Assignment Group field based on the incident's Short Description

Incident Categorization	Classification	Predicts the Category field based on the incident's Short Description.
Major Incident Detection	Similarity	- Recommends similar active Major Incidents which the current Incident can be linked to.

- Recommends similar Incidents to propose a Major Incident.

Similar Incidents (Major Incident Workbench)	Similarity	Recommends similar Incidents that are not linked as child incidents to a Major Incident.
Similar Incidents	Similarity	Recommends similar Incidents to help with incident investigation and resolution process.
Similar Open Incidents	Similarity	Recommends similar open Incidents to help with Incident investigation and resolution process.
Similar Resolved Incidents	Similarity	Recommends similar resolved Incidents to help with Incident investigation and resolution process.

Knowledge Management Knowledge Similar Articles Similarity Suggests Related Articles based on Short Description, in two places: - Knowledge Results section on Knowledge Form when creating a Knowledge Article. - Related Articles section on article view page in the Service Portal and Mobile applications. HR Service Delivery HR Case Categorization Classification Predicts the HR Service from the Short Description and Description fields. Customer Service Management (CSM) Case Assignment Classification Predict the Assignment Group based on the Short Description.

Case Categorization	Classification	Predict the Category based on the Short Description.
Case Prioritization	Classification	Predict the Priority based on the Short Description.
All Similar Cases	Similarity	Recommends similar cases based on the Short Description.
Recommended Open Cases	Similarity	Recommends similar open Cases based on the Short Description.
Recommended Resolved Cases	Similarity	Recommends similar resolved Cases based on the Short Description.
Major Issue Detector	Similarity	Recommends Major Issues based on Short Description.  Also recommends one or more Major Cases, if available; otherwise recommends similar Cases that are not linked as child cases to a Major Case.

Event Management Closed Alert Similarity Similarity Recommends similar Alert records based on: Description, Node, Resource, CI, Source, Metric name. Results visible on the Alert form and on the Similar Alerts tab in Workspace.

- Solution Types: - Classification Solution

1. ?

- Class

1. An output field value that a predictive model uses to make predictions through classification solutions.
2. Each Class is an output field value, along with a list of possible precision, coverage, and distribution metrics to choose from.
3. Class Distribution
4. The percentage of records from the entire table that has a particular output field value.
5. Classification Solutions use Business Rules to set field values.
6. Clustering Solution
7. Example Use Case: Group similar incidents that have occurred together to identify a major incident.
8. Steps to create:

1) Choose a Word Corpus

			2) Specify the table and fields you want to cluster as well as how they should be grouped.
			3) Define the target solution Coverage, as well as the minimum number of records per cluster.
			4) Train the Solution.
		- You can view the results of the clustering operation after you execute the Training operation.
		- A Cluster Visualization scatter plot chart shows the top 50 clusters and their records.
	- Similarity Solution
		Steps to create:
			1) Choose a Word Corpus
			2) Select 2 tables to compare.
			3) Select a re-training frequency and refresh frequency.
				- A refresh frequency allows new records since the last refresh to be added to the Similarity comparison.
			4) Train the Solution.
		- You can review the results of each comparison that has taken place.
		- You can also test the solution on arbitrary input directly on the Test Solution tab by hitting the Run Test button.

- Solution Template

1. A predefined plan for populating fields on the solution definition with predefined values. (Predefined values??)

- Training

1. The act of processing given information, based on the parameters within the Solution Definition, to build a Predictive Model or Word Corpus.
2. Training Frequency - How often the Solution is refreshed/retrained; that is, how often new data is brought in.
3. This is done after you have configured your Solution Definitions.
4. Your data is packaged and sent to a Training Server.
5. Training duration is based on the number of records PI must analyze.

- Training allows a Model to see the “correct answers”, because you are pointing it to completed, correct, records. This allows it to figure out a method for predicting missing data on future records. - Training can take a few hours. - All artifacts (solution definition and historical records) are packaged into an update set (CSV files) and exported to the nearest Training Server at a ServiceNow data center. When training is complete, the server removes all customer data and sends the Trained Model back to the customer instance. - When a solution definition is scheduled to train on an interval, the new solution automatically goes live when the results are returned. Older versions are set to inactive, but can be reactivated later if needed. - Word Corpus - A collection of words and phrases that functions as the vocabulary (dictionary) the system uses to compare records based on their textual similarity. - Provide context to various words used in records. - Necessary before creating a Similarity Solution. Also used in Clustering solutions. - To build a Word Corpus, select the Tables, input fields, and data sets (using filters).

- Principles / Geoff's Principles - Visibility - Visibility into anything you are working with is always very important. Examples:

1. Scope of Work.
2. Objects available to you in code/script. (Do whatever is necessary to Gain Visibility into what is happening!)
3. Debugging tools available.
4. Network traffic.
5. Number of users.
6. Frequency of use.
7. Organizational Structure, Coworkers, and Contact Information.
8. Salaries / Rates

- It's so important to SEE what is going on, and what context you are in when writing or maintaining code.

1. De-Duplication
   1. Keep data in One Place only, so that you do not have synchronization problems.
   2. This also makes it much easier to understand what the source of truth is for data.
   3. It also helps you conceptually, because there are fewer objects (columns, fields) to keep in your head.
   4. Also, try to have only ONE way to Perform each action in the system. Lowers maintenance. Simplifies learning / training.

This means sitting down and typing lines of code, testing, modifying, and so on, until you have built something that works, in the traditional manner. It's like baking from scratch. It may be a good approach when you are building something very new or innovative. It's generally expensive and time-consuming.

- Allows you to group multiple incidents together and find the root-cause. You can also link a Problem to a CI or Business Service inside of a Change Request. A Dependency Views map shows upstream and downstream dependencies for a CI. A typical approach to analysis would be to create a task for each team responsible for the dependent objects. Communicate Workaround sends workaround instructions to all those users who reported incidents attached to the problem record. If a problem is linked to a Change Request, and the change request is completed successfully, the problem and all associated incidents will also be marked as Resolved.

Process Automation - Enterprise Level Product Bundle Required! - Part of ITSM. Enterprise Level. - Based on the Sys Log. - Can operate on anything that is based on the TASK table. - Accessed via the “Analyst Workbench”? - Shows how many tickets have historically moved from one state to another.

- A potential project starts with an Idea. The Idea is managed in the Ideation module. It is vetted and promoted into a Demand, where the business case is built out. Demands are prioritized and after approval a Demand becomes a full-fledged Project. - Project Controls include Risks, Issues, Changes, and Status Reporting. - The Project Management app tracks Tasks, Milestone, and Timelines, with the goal of completing projects on time and within budget. - The Project Management and Resource Management apps give you a comprehensive view of project progression and where resources are being utilized. - Resource Management - The Resource Management App provides Transparency into requested and allocated resources and forecasted and actual Labor Costs.

Provides a simplified, team-oriented approach to Project Portfolio Management and IT development. It includes the following individual applications: - Demand Management - An application used for gathering and assessing ideas and promoting accepted ideas to strategic and operational demands. - Project Management - A suite of tools used to manage projects, task, and resources. - Software Development Life Cycle (SDLC) - An application used for managing the software development and release process. - Test Management - An application that provides tools for manual software testing. - Resource Management - An application that enables resource requesters to create resource plans and request resources.

- Used to safeguard intellectual property by making logic read-only or invisible. - Only applies when an application is installed from the ServiceNow App Store. - May be applied to UI Actions or Script Includes. - “Protected” means Not Viewable.

- Generally everything is available via the Platform UI; and Studio is NOT required.

1. Geoff: Code Search may require Studio though.

- May need Studio for Source Control?? - Roles can contain other roles. hasRoleExactly means that they have the role assigned Directly, not via inheritance. - The “maint” role is used for maintenance, and is only available to ServiceNow employees.