What is an AI Co-Pilot?

An AI Co-Pilot is an artificial intelligence assistant designed to help users perform tasks more efficiently by providing contextual suggestions, automating repetitive actions, and enhancing productivity within existing workflows. It typically works alongside users, offering real-time support without taking independent action.

However, more recently, AI Agents or Agentic AI have arrived as the new kid on the block who seems to be drawing all the limelight (and for good reason!).

What is Agentic AI?  

Agentic AI refers to autonomous AI systems capable of not only assisting users but also acting independently to complete multi-step goals without constant human intervention. These systems leverage advanced automation, conversational interfaces, and decision-making capabilities to deliver end-to-end task execution.

Tosca is evolving with AI!

Tricentis has introduced Agentic AI into Tosca, taking automation to the next level.

While Tosca Copilot acts as an assistive AI—helping with queries, test summaries, and troubleshooting—Agentic AI goes further by autonomously creating, adapting, and executing tests from natural language.

Key Highlights:

Copilot = Assistive AI → boosts usability & guidance

Agentic AI = Autonomous AI → drives end-to-end test automation

Reported 85% faster test creation & 60% productivity gains

This marks a big shift in how QA teams can leverage AI to accelerate testing and improve software delivery.

QA Terminology

Below are the most commonly used QA terminologies that every QA professional ought to know. Make sure you’re familiar with all of them:

1. Test Case: A set of actions used to determine if a system behaves as expected in a given scenario.
2. Bug/Defect: A problem or error that hinders a software program from functioning as expected.
3. Smoke Testing: Basic tests to check if a new software version is stable enough for more in-depth testing.
4. Regression Testing: Verifying if previously working functionality still works after new changes.
5. Unit Testing: Testing the smallest testable parts of an application in isolation (e.g., functionality like a button, link, dropdown, etc). If an application were a physical machine, it would be like testing the quality of the nuts, bolts and transistors before assembled into a more recognizable part of the machine, for instance a control panel. This testing is commonly done by developers.
6. Integration Testing: Testing in which individual components or units are added one by one and tested progressively until the group of components are successfully tested. If an application were a physical machine, integration testing would be like testing how the previously tested nuts and bolts work together once assembled. For example, a control panel, which is only a part of a machine, must have many components or units that must be assembled to test that the panel itself works as it should.
7. System Testing: Testing the integrated parts together to ensure it meets the requirements. If an application were a machine, in addition to a control panel, there would be other machine parts similarly assembled and tested. System testing is like seeing whether the entire machine works as expected.
8. UAT (User Acceptance Testing): The final testing phase where actual users try the software to make sure it works in real-life scenarios. The users can be individual people or a company for which the product/software was created.
9. Test Suite: A collection of test cases that have been grouped for a specific purpose.
10. Sanity Testing: Testing that uses a subset of regression tests to quickly check that a new software build/version works as expected. Some people consider this the same as smoke testing.
11. Black Box Testing: Testing software based on output, without knowing its internal workings. For example, when an end-user interacts with the website’s UI without having access to the code.
12. White Box Testing: Testing software with knowledge of its internal workings.
13. Test Plan: A detailed document outlining the testing strategy, objectives, resources, schedule, and deliverables.
14. Test Script: Step-by-step instructions for a particular test.
15. Test Scenario: A high-level idea of what to test. It can have multiple test cases.
16. Exploratory Testing: Testing the software without a set plan, exploring and learning the application.
17. Boundary Testing: Testing the limits (edges or boundaries) of the software input.
18. Functional Testing: Testing that software features work as expected. This is an umbrella term under which several of the other types of testing already defined are grouped (i.e., Sanity, Smoke, Regression, UAT).
19. Non-functional Testing: Testing non-operational aspects of a software, like performance, usability, or security.
20. Test Environment: A controlled setting where testing is conducted.

What is Tosca Copilot ?

Tosca Copilot is a generative AI-powered assistant integrated into Tricentis Tosca, designed to enhance user productivity in testing. It utilizes advanced large language models (LLMs) to help users quickly find, understand, and optimize test assets. Tosca Copilot facilitates tasks like explaining test cases, converting natural language to Tosca Query Language (TQL), and summarizing test execution results, ultimately aiming to boost efficiency and quality across the testing lifecycle. 

Tosca Copilot is part of the broader Tricentis Copilot program, which includes AI-powered assistants for other Tricentis products like Testim and qTest. This program aims to accelerate test creation, improve test quality, and simplify testing across the entire software development lifecycle.

Here's how Tosca Copilot can help you:

• Test Case Generation: Automatically create test cases from natural language descriptions or user stories, including edge cases and variations, accelerating the test design process.
• Test Optimization: Streamline your test suite by identifying and suggesting removal of unused, duplicate, or unlinked test cases, or by recommending optimizations to existing ones for better coverage and efficiency.
• Test Result Insights: Understand failed tests faster with actionable insights generated by Tosca Copilot, aiding in quick troubleshooting and issue resolution.
• Maintenance and Refactoring: Perform maintenance tasks like renaming test steps, cleaning up labels, or searching for specific test artifacts using chat commands.
• Onboarding and Learning: New team members can quickly grasp complex test cases and Tosca functionalities through explanations provided by the Copilot, according to Tricentis. 

How it works:

Tosca Copilot leverages advanced Large Language Models (LLMs) to understand natural language requests and interact with Tosca, performing actions like: 

• Converting natural language instructions into Tosca Query Language (TQL) queries for finding and managing test assets.
• Generating test case descriptions, steps, and data from user stories or specifications.
• Analyzing execution logs and providing insights into test failures in easily understandable language. 

Benefits:

• Time Savings: Reduced manual effort for creating tests, generating data, and analyzing results translates to faster testing cycles and quicker time-to-market.
• Increased Productivity: Testers can focus on more strategic tasks, and new team members can get up to speed faster with the Copilot's assistance.
• Cost Savings: By optimizing test suites, reducing redundant efforts, and minimizing manual maintenance, Tosca Copilot helps lower overall testing costs.
• Improved Software Quality: Automated test generation and defect analysis lead to higher test coverage, early bug detection, and ultimately, better quality software. 

In conclusion, Tosca Copilot helps streamline test automation by assisting with various tasks throughout the testing lifecycle. It utilizes generative AI to enhance productivity, accelerate learning, reduce costs, and ultimately deliver higher quality software.

Tosca Copilot's capabilities:

Natural Language Interaction:

Tosca Copilot allows users to interact with Tosca using natural language, making it easier to query and understand test assets. 

Test Case Explanation:

It can explain the functionality of a test case in plain language, providing insights into its purpose and steps. 

TQL Query Generation:

Users can convert their natural language queries into Tosca Query Language (TQL) to search and filter test assets effectively. 

Test Optimization:

Tosca Copilot helps in identifying unused or unlinked test assets, duplicates, and other inefficiencies, enabling users to streamline their test libraries. 

Troubleshooting:

It aids in understanding failed test executions by summarizing the results and providing insights into potential causes. 

Integration with Microsoft Azure OpenAI Service:

Tosca Copilot leverages the power of Microsoft Azure OpenAI Service, ensuring enterprise-level data privacy and security compliance. 

How to View Regression Status in 1 Click ?

We have to create BAT file and double click on file We can  check Pass/Fail status in just one click — no Commander needed. 

Creating BAT file: 

🔧 𝐒𝐭𝐞𝐩 𝟏 – 𝐂𝐫𝐞𝐚𝐭𝐞 𝐚 𝐓𝐂𝐒 𝐟𝐢𝐥𝐞 𝐆𝐞𝐭𝐑𝐞𝐬𝐮𝐥𝐭𝐬.𝐓𝐂𝐒

𝑔𝑒𝑡 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑇𝑒𝑠𝑡𝐶𝑎𝑠𝑒𝑠𝑃𝑎𝑠𝑠𝑒𝑑
𝑔𝑒𝑡 𝑁𝑢𝑚𝑏𝑒𝑟𝑂𝑓𝑇𝑒𝑠𝑡𝐶𝑎𝑠𝑒𝑠𝐹𝑎𝑖𝑙𝑒𝑑

🔧 𝐒𝐭𝐞𝐩 𝟐 – 𝐂𝐫𝐞𝐚𝐭𝐞 𝐚𝐧𝐨𝐭𝐡𝐞𝐫 𝐟𝐢𝐥𝐞 𝐑𝐮𝐧𝐓𝐐𝐋𝐓𝐨𝐆𝐞𝐭𝐓𝐡𝐞𝐒𝐭𝐚𝐭𝐮𝐬.𝐓𝐂𝐒

This one jumps to Execution List and calls the result fetcher:

𝐽𝑢𝑚𝑝𝑇𝑜𝑁𝑜𝑑𝑒 "/𝐸𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛/𝐸2𝐸 𝐸𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛/𝐸𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛𝐿𝑖𝑠𝑡𝑠"
𝐹𝑜𝑟 =>𝑆𝑈𝐵𝑃𝐴𝑅𝑇𝑆:𝐸𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛𝐿𝑖𝑠𝑡 𝐶𝑎𝑙𝑙𝑂𝑛𝐸𝑎𝑐ℎ "𝐷:\\𝑇𝑜𝑠𝑐𝑎\\𝑇𝑐𝑆ℎ𝑒𝑙𝑙\\𝐺𝑒𝑡𝑃𝑎𝑠𝑠𝐹𝑎𝑖𝑙𝑅𝑒𝑠𝑢𝑙𝑡\\𝐺𝑒𝑡𝑅𝑒𝑠𝑢𝑙𝑡.𝑡𝑐𝑠"
𝐸𝑥𝑖𝑡

🔧 𝐒𝐭𝐞𝐩 𝟑 – 𝐂𝐫𝐞𝐚𝐭𝐞 𝐒𝐜𝐡𝐞𝐝𝐮𝐥𝐞𝐁𝐚𝐭𝐜𝐡𝐅𝐢𝐥𝐞.𝐁𝐚𝐭

This will open the workspace, authenticate, and call the above script:

"𝐶:\𝑃𝑟𝑜𝑔𝑟𝑎𝑚 𝐹𝑖𝑙𝑒𝑠 (𝑥86)\𝑇𝑅𝐼𝐶𝐸𝑁𝑇𝐼𝑆\𝑇𝑜𝑠𝑐𝑎 𝑇𝑒𝑠𝑡𝑠𝑢𝑖𝑡𝑒\𝑇𝑜𝑠𝑐𝑎𝐶𝑜𝑚𝑚𝑎𝑛𝑑𝑒𝑟\𝑇𝐶𝑆ℎ𝑒𝑙𝑙.𝑒𝑥𝑒" -𝑤𝑜𝑟𝑘𝑠𝑝𝑎𝑐𝑒 "{𝑊𝑜𝑟𝑘𝑠𝑝𝑎𝑐𝑒𝑃𝑎𝑡ℎ}" -𝑙𝑜𝑔𝑖𝑛 "{𝑈𝑠𝑒𝑟𝑁𝑎𝑚𝑒}" "{𝑃𝑎𝑠𝑠𝑤𝑜𝑟𝑑}" "{𝑅𝑢𝑛𝑇𝑄𝐿𝑇𝑜𝐺𝑒𝑡𝑇ℎ𝑒𝑆𝑡𝑎𝑡𝑢𝑠.𝑇𝐶𝑆 𝐹𝑖𝑙𝑒 𝑃𝑎𝑡ℎ}"

🔧 𝐒𝐭𝐞𝐩 𝟒 – 𝐂𝐫𝐞𝐚𝐭𝐞 𝐨𝐧𝐞 𝐟𝐢𝐧𝐚𝐥 𝐁𝐀𝐓 𝐟𝐢𝐥𝐞 𝐭𝐨 𝐬𝐭𝐨𝐫𝐞 𝐨𝐮𝐭𝐩𝐮𝐭 𝐢𝐧 𝐚 .𝐭𝐱𝐭 𝐟𝐢𝐥𝐞

"{𝑆𝑐ℎ𝑒𝑑𝑢𝑙𝑒𝐵𝑎𝑡𝑐ℎ𝐹𝑖𝑙𝑒.𝐵𝑎𝑡 𝑓𝑖𝑙𝑒 𝑝𝑎𝑡ℎ}" > "{𝑌𝑜𝑢𝑟𝑅𝑒𝑠𝑢𝑙𝑡𝐹𝑜𝑙𝑑𝑒𝑟𝑃𝑎𝑡ℎ\𝑇𝑒𝑠𝑡𝑅𝑒𝑠𝑢𝑙𝑡.𝑡𝑥𝑡}"

✅ End Result?

You get a TestResult.txt file with Pass/Fail counts ✔️
→ Quick
→ Clean
→ Commander-free

🔁 𝐁𝐨𝐧𝐮𝐬 𝐓𝐢𝐩: 𝐒𝐜𝐡𝐞𝐝𝐮𝐥𝐞 𝐭𝐡𝐢𝐬 𝐁𝐀𝐓 𝐟𝐢𝐥𝐞 𝐰𝐢𝐭𝐡 𝐖𝐢𝐧𝐝𝐨𝐰𝐬 𝐓𝐚𝐬𝐤 𝐒𝐜𝐡𝐞𝐝𝐮𝐥𝐞𝐫 — 𝐠𝐞𝐭 𝐲𝐨𝐮𝐫 𝐫𝐞𝐩𝐨𝐫𝐭 𝐫𝐞𝐚𝐝𝐲 𝐛𝐞𝐟𝐨𝐫𝐞 𝐲𝐨𝐮𝐫 𝐜𝐨𝐟𝐟𝐞𝐞 𝐛𝐫𝐞𝐰𝐬 ☕

REM Example BAT file content
cd "C:\Program Files\TRICENTIS\Tosca Commander"
TCShell.exe -f "C:\Path\To\Your\Script.tcs" -w "C:\Path\To\Your\Workspace.tws" -u "YourUsername"

-p "YourPassword"

What is testRigor ?

 testRigor is an AI Agent that allows anyone to create end-to-end tests from an end user's perspective using plain English, therefore eliminating excessive test maintenance related to locator changes. testRigor supports testing on the following platforms:

• Web testing (Windows, MacOS, Ubuntu) and Mobile Web testing on iOS and Android
• Native and Hybrid Mobile App testing for iOS and Android
• Native Desktop applications testing
• Mainframe application testing

With testRigor, you can perform various types of testing, including:

• Acceptance testing
• Smoke testing
• Regression testing
• System (end-to-end) testing
• API testing
• Visual testing
• SMS and phone call testing
• 2FA and Captcha testing

To create your end-to-end tests, you have several options:

• Leverage testRigor's Generative AI to create tests based on descriptions
• Write tests from scratch using plain English commands (See this documentation for help)

testRigor

 What is testRigor ?

What is Execution Recording in Tosca ?

--When you run a test in Tosca, Execution Recordingautomatically keeps track of:
-What steps were performed
-What data was used
-What passed or failed
Benefit-This helps in finding issues, showing proof of testing, and creating reports.
🛠️ How Does It Work?
1. Turn on Execution Recording
* You can switch it on in the Tosca settings.
* Path: Project → Settings → Execution → ExecutionLogEnabled = True
2. Run Your Test
Run your test from Tosca Commander*or an ExecutionList.
* Tosca will record:
-Every step clicked or typed
-Input/output data
-Results of each step (Pass/Fail)
-Screenshots (if enabled)
3. See the Results
* Go to the ExecutionList → ActualLog.
* You can get below:
-Which steps ran
-Which ones passed or failed
-Any error messages
-Screenshots (if set up)
4. Detailed View
--Click on any test step to check:
-What was expected
-What was actually done
-What data was used
5. Export or Share
You can save the results as:
-PDF
-Excel
-HTML
You can send them to tools like: qTest,JIRA,Jenkins.