AI Agents & Codeless Automation

 Limitations of LLM's

  1. Can't perform real time operations
  2. no memory or context of previous conversations
  3. Can't perform live operations like connecting to database etc.
  4. Task automation not possible.
  5. Results might be biased due to dataset that we are training
  6. Not upto date with data, most of the LLM's are trained till 2022, they are not real time

AI Agents

  1. Above all limitations can be addressed using AI agent, we can book train tickets, flight, even ask to create ticket by providing information etc.
  2. test rigor is a tool which helps in doing it (Generative AI based test automation tool)
    1. add test case / edit test case
    2. parameterize data from test data like stored value "email"
      1. make sure create email in test data
    3. resusable rules (sub routines) you can add resusable code like login, logout etc


Comments

Post a Comment

Popular posts from this blog

Evaluate RAG - LLM's using RAGAS Python Pytest Framework

Image
 Data will be fed to embedded model which stores the data into vector database based on semantic understanding. LLM(like chatgpt, gemini, anthropic) interacts with this vector database and replies based on the user prompts RAG (Retrieval Augmented Generation) data retrieved from vector database articulation of whole question with necessary context is called augmentation Generation is about giving the answer In traditional software testing, we assert on the system responses, but in LLM's we evaluate quality of the response using benchmark metrics. Here we make use of RAGAS framework Benchmark metrics for document retrieval system context precision --> no of relavant docs retrieved / total number of docs retrieved => 3/6 50% context recall =>  no of relavent docs retrieved / total number of relavent docs retrieved Install Python Install Pycharm IDE make sure to configure path of python while importing project (use which python3 to know the path of installation  Let...
Read more

Spring Microservices

Objective: To develop micro services using Spring. We will also look at how to containarize it with Docker and orchestrate via kubernates. What is a Web Service? In layman terms, Service delivered over the web. 1. designed to communicate from one application to application 2. not platform dependent (interoperable) 3. should allow communication over a network. Service Definition defines structure of request and response and other details Data exchange takes place using xml or json. Key terms: 1. Request => input to a web service 2. Response => output from a web service 3. Message Exchange Format => XML or JSON 4. Service Provider - which hosts the web service 5. Service Consumer - which is consuming web service 6. Service Definition - contract b/w service provider and service customer 7. Transport - How service is called like HTTP or MQ - web sphere MQ ( more like a queue ) - I mean communication happens over a queue. Web service Groups 1. SOAP - xml request/response   ...
Read more

Appium 7 - Android Test Automation

Image
Agenda: To understand Appium and explore concepts that helps us to achieve mobile test automation. Introduction: Open source tool that works for native, hybrid and mobile apps Supports selenium webdriver API Platform support for iOS, Android and FirefoxOS Architecture: Appium Client Code => Json (Request) ==> Appium Server ==> UIAutomator2 ==> Android                                                                                                  ==> XCUITest        ==> iOS Appium client code is the code that is written by user using any language that is supported by selenium like Java, C#, python etc The request will be sent to Appium server in the form of JSON, which is interpreted by server and executes ac...
Read more