Curated interview practice questions with answers. These are practice scenarios, not a verified list of past interview questions.
About HSBC hiring for Automation Testing roles
HSBC is a global banking group with retail, commercial and wholesale banking operations across Asia, Europe and other regions. Its India technology and global service centres support software development, operations, analytics and risk for the group. It hires for software engineers, testers, data analysts, data engineers and financial crime analysts.
For Automation Testing roles, this HSBC practice set focuses most on Selenium WebDriver (8), API Testing and CI/CD (5), Cucumber and BDD (4), Selenium Grid & Docker (4). It has 7 easy, 29 medium, 4 hard questions, so you can start easy and work up to the harder ones.
Company type
Banking & finance
Headquarters
London, United Kingdom
Main India offices
Hyderabad, Pune, Bengaluru
HSBC Automation Testing interview process
Online assessment (aptitude + technical)
Technical round
Managerial round
HR round
Typical pattern based on public candidate experiences. Rounds vary by role, team and year.
Solved for HSBCStart practicing
3 QUESTIONS FREESign up to unlock the full question bank and track your progress.Unlock every question and answer, then continue practising company-wise and package-wise.
Implicit wait applies globally to element lookups. Explicit wait waits for a specific condition (WebDriverWait + ExpectedConditions). Fluent wait is an explicit wait with custom polling and ignored exceptions.
02
What are pre-processors and post-processors in JMeter?
Pre-processors run before a sampler, for example a JSR223 PreProcessor that builds a signature or timestamp, or User Parameters that set values per thread. Post-processors run after a sampler and usually extract data from the response, such as JSON Extractor, Regular Expression Extractor or Boundary Extractor, storing it as a variable for later requests. Their scope depends on where they sit in the tree, so place them as children of the right sampler.
03
Why is JSR223 with Groovy preferred over BeanShell in JMeter?
JSR223 elements with Groovy compile scripts once and cache them, so they run much faster and use less CPU under load. BeanShell is interpreted on every call and becomes a bottleneck at high thread counts. Groovy also supports modern Java syntax and is the default JSR223 language. Enable caching by keeping the cache compiled script option on and passing values through vars or parameters, not by embedding ${} inside the script.
04
Tell me about a time your API tests caught a breaking change before the UI tests did.
Situation: a backend team renamed a response field during a refactor. Task: our nightly REST Assured suite had to protect the mobile and web clients. Action: schema validation failed within minutes of the merge, and I shared the failing response and the contract with the developer. Result: the field was restored with a deprecation plan before the UI suite even ran, and the team added our schema tests to their pull request checks.
Add tags like @smoke or @regression above a Feature or Scenario, then filter with a tag expression. In the runner use @ConfigurationParameter(key = FILTER_TAGS_PROPERTY_NAME, value = "@smoke and not @wip") with the JUnit Platform engine, or pass -Dcucumber.filter.tags="@smoke" from Maven. Tag expressions support and, or, not and brackets, so CI can run smoke on every commit and regression nightly.
String parent = driver.getWindowHandle();
for (String h : driver.getWindowHandles())
if (!h.equals(parent)) driver.switchTo().window(h);
// ...then switch back to parent
On a configuration method like @AfterMethod(alwaysRun = true), it makes the method run even when tests fail or are skipped, or when groups filter would exclude it. It is essential for cleanup such as driver.quit(). On a test with dependsOnMethods, alwaysRun = true runs it even if the dependency failed. Missing alwaysRun on teardown is a common cause of leftover browsers in CI.
09
Write a docker-compose.yml that runs a Selenium hub with Chrome and Firefox nodes.
Situation: leadership set a goal to automate every manual test case. Task: set realistic expectations. Action: I analysed the cases, showed that some were one-time, visual or exploratory, and proposed covering critical and repetitive flows first with a clear ROI estimate. Result: we agreed on 80 percent of regression cases, delivered it on time, and kept exploratory testing manual.
11
What is the difference between an array and an ArrayList?
They are test doubles that replace real dependencies. A stub returns fixed answers, such as a fake API returning a canned user. A mock also verifies how it was called, like checking a payment service received one charge. A fake is a lightweight working version, such as an in-memory database. A spy wraps a real object and records calls. Automation testers often use WireMock stubs to isolate tests from unstable services.
13
Tell me about a time you worked with a team in a different location or time zone on automation.
Situation: our developers were in another time zone and test failures waited a day for replies. Task: reduce delays. Action: I wrote clear failure reports with screenshots, logs and steps, recorded short videos for complex issues, and set a fixed 30-minute overlap call twice a week. Result: turnaround on defects dropped from two days to about one, and fewer issues bounced back.
14
How do you undo a commit that has already been pushed?
Use git revert <commit>, which creates a new commit that reverses the changes and keeps history intact, safe for shared branches. git reset moves the branch pointer backwards and is fine for local commits, but resetting and force-pushing a shared branch rewrites history and breaks teammates' clones. If force push is truly needed, use --force-with-lease and tell the team.
15
Write Selenium code to verify that all links on a page are not broken.
Retries are hiding real intermittent defects, such as race conditions, by treating them as flaky tests. Limit retries to one, report retried tests separately as flaky in the dashboard, and require a ticket for each one. Review retried tests weekly to find whether the cause is the test or the product. Retries should buy time to fix, not replace fixing.
Encapsulation: private locators with public page methods. Inheritance: pages extend a BasePage. Polymorphism: the WebDriver interface for different browsers. Abstraction: utility layers hiding Selenium details.
18
What are timers in JMeter and why is think time important?
By default JMeter fires requests back to back, which is far more aggressive than real users who read pages and fill forms. Timers add delays: Constant Timer, Uniform Random Timer and Gaussian Random Timer add think time, while Constant Throughput Timer and Precise Throughput Timer control request rate. Realistic think time makes the user count meaningful; without it, 100 threads may generate the load of thousands of real users.
git cherry-pick <commit> applies the changes of one specific commit onto your current branch, creating a new commit. It is useful when a fix on one branch, such as a locator fix on develop, is urgently needed on a release branch without merging everything else. Overuse creates duplicate commits and confusing history, so prefer normal merges when possible.
20
How do you attach a screenshot to the Cucumber report when a scenario fails?
GraphQL usually has one endpoint and you send a POST with a query and variables. Tests check the data returned matches the requested fields, that errors appear in the errors array, often with status 200, and that authorization applies per field. Also test query depth limits and nested queries for performance. REST Assured works fine by sending the query as JSON body.
23
How do you make page objects work with dynamic locators, such as a row containing a given name?
Write a method that builds the locator from a parameter, for example private By rowFor(String name) { return By.xpath("//tr[td[normalize-space()='" + name + "']]"); }, and expose actions like deleteUser(name). Keep the template in the page class, not in tests. Escape quotes if names may contain them, and prefer CSS with data attributes when the application provides them.
Low-end devices are slower, so fixed waits and short timeouts expire before screens render. Replace sleeps with explicit waits and increase timeouts sensibly. Check for vendor-specific pop-ups, battery optimisation dialogs or different keyboards on that brand. Compare screen size and resolution, since coordinate-based taps break. Collect device logs with adb logcat and screen recordings to see what the app was doing. Also check whether the app crashed due to low memory.
26
What is the difference between a page object and a utility class in a test framework?
A page object models one screen or component: its locators and the user actions possible there, such as LoginPage.loginAs. A utility class holds generic helpers not tied to any page, such as reading config, generating test data, date formatting or taking screenshots. Mixing them makes page objects bloated and utilities dependent on the UI. Keep page objects in a pages package and helpers in a utils package.
27
How do you start a standalone Chrome container and run your tests against it?
Run docker run -d -p 4444:4444 -p 7900:7900 --shm-size=2g selenium/standalone-chrome:latest, preferably with a pinned version tag instead of latest. It starts a mini Grid with one Chrome node. Point RemoteWebDriver to http://localhost:4444 with ChromeOptions and the tests run inside the container. Port 7900 lets you watch through noVNC. Stop it with docker stop when finished, or use --rm so it is removed automatically.
28
What is the LoadableComponent pattern in Selenium?
LoadableComponent is a Selenium support class that standardises how a page object ensures it is loaded. You extend it and implement load(), which navigates to the page, and isLoaded(), which throws an Error if the page is not ready. Calling get() checks isLoaded and calls load only if needed, then checks again. It gives a consistent, self-verifying way to open pages and fails fast with a clear message.
First check timing: the screen may still be loading, so add an explicit wait for presence. Then check context: in a hybrid app the element may be in a webview while you are in NATIVE_APP. Confirm the locator matches exactly, including package prefix in resource-id. The element may be off screen in a list, so scroll first. Also check that the Inspector session used the same app build, language and device size as the test.
32
How do you store passwords and tokens safely in Jenkins?
Use the Jenkins Credentials store and bind them in the pipeline with withCredentials or the credentials() helper in environment blocks. Jenkins masks the values in console logs. Never hard-code secrets in the Jenkinsfile, pom.xml or properties files committed to Git. For larger setups, integrate a vault service and give each job only the credentials it needs.
33
What are the standalone, hub and node, and distributed modes of Selenium Grid 4?
Standalone runs every Grid component and a local node in a single process, which suits a laptop or a single container. Hub and node mode runs the router, distributor, queue and session map together as the hub, with separate node processes registering to it, which is the usual team setup. Fully distributed mode runs each component separately, often on Kubernetes, for very large grids that need independent scaling and high availability.
34
Tell me about a time you found a critical bug through automation.
Situation: during a sprint, my nightly regression showed checkout failing only for one payment option. Task: confirm if it was a test issue or a real bug. Action: I reproduced it manually, checked the API response and found a wrong tax value, then logged a clear defect with steps, logs and screenshots. Result: developers fixed it before the release and the test was added to the smoke suite.
35
Tests running inside Docker cannot reach the application running on the host's localhost. Why, and how do you fix it?
Inside a container, localhost means the container itself, not the host machine. On Docker Desktop use host.docker.internal as the host name. On Linux add extra_hosts with host.docker.internal:host-gateway, or use the host's IP. A cleaner option is running the application in the same Compose network and calling it by its service name, such as http://web:8080. Make sure the app listens on 0.0.0.0, not only 127.0.0.1.
36
What is a DataTable in Cucumber and how do you use it?
A DataTable passes a table written under a step into the step definition. You can read it as List<String>, List<Map<String,String>>, Map<String,String> or convert rows into objects with a @DataTableType method. For example a step that creates several users can take a table of name, role and email. Unlike Scenario Outline, a DataTable feeds data into one step and the scenario runs once.
A callback where one system sends an HTTP request to another when an event occurs. Test with a mock receiver or tools such as webhook.site and verify the payload.
A quality gate is a rule a build must pass before moving forward, such as all smoke tests passing, code coverage above a threshold, no critical static analysis issues, or no new high-severity vulnerabilities. If the gate fails, the pipeline stops and the change is not deployed. Gates make quality an automatic decision instead of a manual check.
39
Should every test in a project be written in Cucumber?
No. Cucumber adds a translation layer, so it is worth it for business-facing acceptance tests that stakeholders discuss. Unit tests, low-level API checks, data validations and technical edge cases are faster to write and maintain directly in TestNG or JUnit. A healthy project uses Cucumber for a focused set of key behaviours and plain code tests for everything else.
40
Why is XPath slow in Appium and what should you use instead?
For every XPath query Appium must fetch the full page source of the app, convert it into an XML document and then evaluate the expression, which can take seconds on large screens and is even slower on iOS. Index-based XPath also breaks when layouts change. Use accessibility id first, then resource-id, Android UiAutomator selectors or iOS predicate strings and class chains, which are evaluated natively on the device and are much faster.
Tips to crack the HSBC Automation Testing interview
Learn anti-money laundering and KYC basics, since financial crime analytics is a large function at HSBC.
Revise SQL, Python and testing fundamentals for technology and data roles in banking systems.
Read about HSBC's focus on Asia and international trade to show understanding of its business.
Keep a one-minute story ready for every project on your CV: problem, your part, tools, result.
Practise the hard questions on this page aloud; HSBC interviewers usually ask follow-ups on your first answer.
Frequently asked questions
How many rounds are there in the HSBC Automation Testing interview?
Candidates usually go through 4 stages: Online assessment (aptitude + technical); Technical round; Managerial round; HR round. The exact rounds change with the role, team, location and your experience, so confirm the process with the recruiter.
Which topics should I prepare for the HSBC Automation Testing interview?
This practice set leans on Selenium WebDriver (8), API Testing and CI/CD (5), Cucumber and BDD (4), Selenium Grid & Docker (4). Revise those first, then practise explaining one or two of your own projects end to end, because most HSBC interviewers ask you to walk through real work.
Are these the exact questions asked at HSBC?
No. These are curated practice questions that match the skills and difficulty usually tested in banking & finance interviews like HSBC's. They are not a leaked or verified list of past questions.
Where does HSBC hire Automation Testing professionals in India?
HSBC has large teams in Hyderabad, Pune, Bengaluru. Openings and locations change often, so check the HSBC careers page and filter by city and experience.
Automation Testing questions for similar companies