QA Engineer (AI-First)

AdVu; Pakistan; On-site; PKR 900,000 - 1,300,000; Full time

Benefits offered by this job

Ownership

Direct access to founder

AI-native product

Competitive pay

Job summary

AdVu is seeking a QA engineer to own end-to-end quality for our AI-powered marketing platform. You will translate specs into acceptance criteria, design test plans, and run both manual and automated tests across the web app and backend.

You will apply AI-assisted techniques to draft test cases, review AI outputs, and contribute to a fast, reliable release cycle while guarding data integrity and user experience across campaigns.

Qualifications

• 2–4 years in software QA for web or SaaS products.
• Solid QA fundamentals: SDLC/STLC, test design, regression, risk-based testing.
• API testing with Postman or code; REST, JSON, token auth.
• SQL to validate data in PostgreSQL.
• Daily use of AI tools in QA, with examples of time saved and AI mistakes caught.
• Proficiency with Jira, Git and CI pipelines; clear bug reporting in English.
• Strong product sense: explain why UX is poor, not just when it’s broken.

Responsibilities

• Turn specs into acceptance criteria and test plans; align with engineering.
• Design test cases and perform exploratory testing on each release.
• Write reproducible bug reports with steps, expectations, evidence, severity.
• Automate end-to-end tests for the React app (Playwright) and API tests (Django).
• Run test suites in GitHub Actions for PRs and deployments.
• Use AI to draft test cases, review AI outputs, and share effective prompts.
• Test AI features (chat, insights, health checks) and guard data & integrations.
• Shape the product by flagging gaps and advocating for user-centric flows.

Skills

QA end-to-end ownership

Test design techniques

API testing (Postman)

SQL (PostgreSQL)

AI tools usage in QA

Jira, Git, CI

Bug reporting & tracking

Product sense

Tools

Postman

GitHub Actions

Jira

Playwright

Job description

Company Description:

AdVu is an AI-powered marketing intelligence platform that helps businesses plan, launch, optimize, and analyze digital marketing campaigns from a single, unified system. By consolidating data and workflows across channels such as Google Ads, Meta, TikTok, programmatic networks, and analytics tools, AdVu reduces complexity and inefficiencies in modern media buying. The platform uses advanced AI and machine learning to turn fragmented marketing data into clear, actionable insights that drive better performance. AdVu enables teams to automate optimization, detect anomalies, and forecast results without relying on multiple disconnected tools. Built for brands, agencies, and growth teams, AdVu’s mission is to turn marketing chaos into clarity through AI and become the operating system for modern media buying.

Role Description:

You’ll own AdVu's QA end-to-end, from reviewing a spec to signing off a release. You’ll use AI to move faster at every step, and you’ll test the AI inside our product, where the same question doesn’t always get the same answer.

We also want your product judgment. When something works but feels wrong to a user, you’ll say so and propose a better way.

What you’ll do

Own QA end to end

• Turn specs into acceptance criteria and test plans, and agree them with engineering before build.
• Design test cases with sound techniques (boundaries, equivalence classes, negative paths), and run
• exploratory sessions on every release.
• Write bug reports developers can act on first time: steps, expected vs actual, evidence and severity.
• Track escaped defects and flaky tests, and fix the process behind them.

Automate what matters

• Build and maintain end-to-end tests for our React web app (Playwright preferred).
• Write API tests for our Django REST backend, covering auth, permissions and error handling.
• Run test suites in GitHub Actions so every pull request and deploy is checked.

Use AI in your daily work

• Use AI assistants to draft test cases from specs, write and refactor test code, generate test data and
• Review AI output before you trust it: you own the result, not the tool.
• Share the prompts and workflows that work, so the whole team gets faster.

Test our AI features

• Test the AI chat, insights and campaign health diagnoses, where answers can vary between runs.
• Extend our golden datasets and evaluation checks for accuracy, consistency and hallucinations.
• Confirm every number the AI quotes matches the source data.

Guard data and integrations

• Cross-check metrics such as spend, CTR, CPA and ROAS against what each ad platform reports.
• Test platform connections end to end: account linking, expired access, delayed data and API limits.
• Verify that each client sees only their own accounts and data.
• Test billing and subscription flows, including failed payments and refunds.

Shape the product

• Review specs and designs early, and flag gaps, edge cases and confusing flows.
• Turn bug patterns and support tickets into concrete improvement proposals.
• Speak for the user in sprint planning and release decisions.

What we’re looking for

Must-haves

• 2–4 years in software QA, at least one of them on a web or SaaS product.
• Solid QA fundamentals: SDLC/STLC, test design techniques, bug lifecycle, regression and risk-based
• testing.
• API testing with Postman or code, and a working grasp of REST, status codes, JSON and token-based
• auth.
• Enough SQL to check data in PostgreSQL yourself.
• Daily use of AI tools (ChatGPT, Claude, Copilot, Cursor or similar) in QA work, with examples of time saved
• and of AI mistakes you caught.
• Comfortable with Jira (or similar), Git and CI pipelines.
• Clear written English: your bug reports need no follow-up questions.
• Product sense: you can explain why something is a poor experience, not just that it’s broken.

Nice to have

• Experience testing LLM or chatbot features, or building evaluation datasets.
• Familiarity with Google Ads or Meta Ads Manager, and metrics like CTR, CPC and ROAS.
• Performance testing (k6, JMeter) or basic security testing (OWASP Top 10, roles and permissions).
• Product-company or startup experience, where you questioned specs instead of only testing against
• them.
• ISTQB Foundation or a similar certification.

What we offer

• Ownership: you set the quality bar for the whole product.
• A real say in what we build, with direct access to the founder and Head of IT.
• Daily work on an AI-native product, with AI built into how we work.
• Competitive pay, benchmarked to your experience and the market.