Software Development Engineer in Test (SDET)

Software Development Engineer in Test (SDET)

Company: Namasys Analytics (Kynaa AI)

Location: India (remote-friendly)

Type: Full-time

Experience: 2 to 5 yrs

About the role

We are building a real-time, voice-driven AI product that conducts live spoken conversations with users and produces AI-scored assessments. You would reporting directly to the product owner, and you would own quality across the entire platform: the web application, the backend services, the live voice experience and the AI behaviour itself.

This is not a conventional SDET seat. Roughly half the product is deterministic software that can be tested with scripts and suites. The other half is a live AI agent whose behaviour is probabilistic, judged through structured evaluation harnesses, transcript review and your own ear.

What you will own

Test automation (greenfield). Backend test suites exist (pytest) and frontend unit tests exist (vitest), but there is no end-to-end suite today. You will design and build our browser automation from scratch; Playwright is our intended direction (we are open to ideas). What you build becomes the release backbone of the platform.

Manual and exploratory testing. Structured regression walkthroughs of the user journeys (the end-user product plus two administrative consoles), device and browser coverage, and bug reports precise enough to act on without a follow-up call.

AI behaviour quality. We maintain evaluation harnesses (Python) that A/B-test changes to the AI interviewer and the AI scoring pipeline against golden baselines. You will learn to run these, then own them: running evals before AI changes ship, auditing real conversation transcripts for behavioural drift, and turning every escaped defect into a new eval fixture.

Release verification. Our deploy pipeline gates every release on automated checks; human verification of what actually shipped is manual. You will start by verifying changes before merge and grow into owning post-deploy verification and release sign-off as you earn context on the production environment.

What we require

• 2 to 5 years in QA or SDET roles, with real automation code you wrote yourself (not just maintained).
• Strong Python. Our backend tests and evaluation harnesses are Python; you must be productive in it from week one.
• Browser automation experience (Playwright preferred, Selenium or Cypress acceptable) and the judgement to know what deserves automation and what does not.
• Fluent spoken English, non-negotiable. Testing this product means taking live voice sessions yourself, daily, speaking aloud and judging conversational quality. You cannot test our core product without this.
• Exploratory testing instinct: you find the bug the test plan did not anticipate.
• Comfort with ambiguity and ownership. You will define the QA function, not inherit one.

What we would love

• TypeScript familiarity (our frontend and future E2E suite).
• Any exposure to LLM-based products, prompt evaluation or non-deterministic testing.
• API testing depth (we run FastAPI) and basic SQL.
• Experience as an early or solo QA hire.

How we work

• Small team, high trust, direct communication. Production is actively developed and releases ship frequently, so quality work has immediate, visible effect. You will handle real user data under strict data-protection rules; discretion and care are part of the job.

Interested Candidates can share their resume at [Confidential Information]