New Technologies - ML, Python, Data Science, Devops : What is in store for Testing Community?

There were days when people used to go in bullock carts, then came bicycles, two / four wheelers, followed by aircrafts etc, there were also times when people used to commute in buses, now we have metro and bullet train is in the making

Contrast above with software testing, soon we may hear the following - there were times when test engineers used to do manual testing (i won't say it will be completely eliminated, but very few technologies and very very few companies may still adapt this testing technique in the future), then we shifted to automation based UI / API testing plus a select few performance test suites (one's that would really help find the bottlenecks).

World is now moving towards Analytics, Machine Learning, Devops

• What is in store for a tester w.r.t these techniques?
• How can a tester (like me and you) learn these new skill sets and yet contribute to the organization in a better way

Python

It is a scripting language, but is being adapted heavily in solving complex analytical problems. Python's easy to use syntax / semantics and libraries it supports makes it a must learn skill set for any tester.

Where can a tester use it? - We write automated tests / perf / security tests, use python as a scripting language

Data Science

Data science enables users to get best search results from a list of millions of results for a specific search keyword (one of the many use cases)

Where can a tester use it?

• When we run large performance test suites, amount of logs generated is enormous, to parse this data takes time, we can use a language like R and get meaningful data, saves time while analysing and creating detailed reports
• Understanding production logs gives a lot of insights, what time a particular feature is being used, peak traffic (higher response times) etc. Tester can also contribute as they would not only improve the product quality but also enhance the regression test scenarios

Devops

Devops as you may know reduces the build and release cycle time. This is where a tester can contribute easily. How? Read on :

• Many of us write automated scripts and integrate to CI CD, we are at one end of the spectrum, we can also delve into setting up the CI CD Job, Add plugins for reporting etc and contribute to the other end of the spectrum too.
• Collaborate with Development, Test, Dev Ops to form a workflow, that on a click of a button deploy, run smoke tests, accept or release a build to next environment and send automatic test reports. This saves huge amount of regression effort manually and gives confidence to stakeholders early in the cycle

Machine Learning (ML)

Suggestions to users in a E-Commerce site / Chat bots answering your queries etc are some of the many use cases that ML has championed. How can a tester use ML?

• This is a vague idea but here is what i feel. Can we build models that tries different exploratory, functional, boundary value, equivalence partitioning etc techniques on a set of a similair applications and come up with all the test scenarios. This reduces writing test cases completely (we can still edit them to customize to our needs) and also get a list of many test scenarios that might have been missed when a individual writes them

Well i am sure there are many more ideas on these new technologies and i would like to hear from you on them "from a tester's perspective"

Be it a Public Vehicle or a Bullet Train, we still need checks in place to ensure system's work as expected, however as they say to do these checks we need to be smarter and not rely only on black box techniques. I am sure testers will master these technologies and break application (then make it) at faster pace and have fun filing those yummy bugs!!!