Arena Entertainment is a digital entertainment company building online platforms and games since 2021, with brands including MetaWin, WOW Vegas, Hit.com, BetZoo Media and Rolla. We run an AI-first engineering culture: we expect the tools to change every few months, and we hire people who are energised by that rather than tired of it.
We build and operate casino games across MetaWin, HIT and our social casino platforms. The role covers three product lines, and you'll work across all of them:
Slots - our own titles such as Legend of Tartarus, Candy Cash and Rise of Cetus: paytables, bonus rounds, free spins, multiplier ladders and RTP-relevant behaviour.
Crypto originals - Plinko, Roulette, Crash and Prime Dice: on-chain settlement, token precision and provably-fair verification.
Live casino - streamed dealer tables: real-time game state, bet windows, stream quality and third-party provider integrations.
These are real-money products. When a payout is wrong, it's wrong in someone's wallet so the bar for correctness is higher than most web QA roles, and the interesting problems live below the UI: settlement maths, wallet reconciliation, provably-fair verification, real-time socket traffic, and the network layer where the truth actually lives.
You'll join a small games QA function in a senior seat, working alongside our existing QA lead who brings deep domain knowledge of our games, our platform and where the bodies are buried. Senior here means you set the standard: you own test strategy for the titles you take, you decide what gets automated and how, and you're the person the team believes when you say something is or isn't ready to ship.
The work is a genuine mix of manual and automated testing, and it stays that way. New builds need someone playing them properly exploratory work is how most of the interesting bugs surface, and live tables in particular have to be sat at and played. Alongside that, we run a Playwright/TypeScript harness with page objects, payout specs and network-intercept probes, with AI-assisted authoring using Claude, and we're actively pushing into agent-driven testing: agents that explore a build, generate and maintain specs, triage failures and cut the regression cost of shipping several titles in parallel.
Coverage is still expanding and the AI side is still being invented. You'll have real influence over both, rather than inheriting someone else's finished framework.
The roadmap runs several titles in parallel. If you like variety, ownership, and finding the bug nobody else would have caught, this is a good fit.
Design and execute test plans for new slot releases, from first playable build through to live production.
Verify win evaluation against the maths spec: symbol combinations, line and ways evaluation, scatter counts, wild substitution checked against the API round response, not the number the reels animate to.
Test bonus and free-spin features end to end: trigger and retrigger conditions, spin counts carrying over, sticky and expanding wilds, feature-buy where offered, and what the feature pays versus what it should.
Exercise multiplier ladders, cascading and tumble wins, progressive and jackpot contributions, and win-cap truncation at the max-win boundary.
Cover bet handling in full: bet levels and coin denominations, min and max stake per currency, autoplay with loss limits and single-win stops, turbo and quick spin, and rapid repeated spins.
Reconcile the wallet on every round balance before and after, insufficient funds mid-autoplay, currency and denomination switching mid-session.
Break the states nobody demos: refresh mid-spin, disconnect mid-bonus, browser back, session resume, network loss during free spins, and a round settled twice.
Validate game history and round replay against what actually happened, including round IDs traceable to the transaction record.
Run long or seeded sessions to sanity-check outcome distribution against expected RTP and volatility, and flag anything that drifts.
Check paytable text against actual behaviour, plus max-win and promotional wording, currency formatting and locale variants.
Test our crypto originals Plinko, Roulette, Crash, Prime Dice where settlement is on-chain and the maths is the product.
Verify provably-fair mechanics end to end: the pre-committed server seed hash, client seed changes, nonce incrementing per bet, seed rotation revealing the previous seed, and outcomes a player could independently reproduce from the published algorithm.
Test each title's own maths: Plinko row counts, risk levels and multiplier grids; Crash cash-out timing, auto-cash-out and multi-bet; Dice roll-over/under targets against payout multiplier and house edge; Roulette inside, outside and neighbour bets against the payout table.
Validate token precision and rounding per currency decimal places, rounding direction, dust amounts, and the difference between the stored value and the displayed one.
Reconcile the ledger: deposits, bets, wins and withdrawals against both the platform wallet and the chain, including pending, confirmed and late confirming transactions.
Hammer concurrency: double-submitted bets, spam betting, the same account open in two tabs or on two devices, and races between a bet and a balance update.
Probe the limits: minimum and maximum stake per currency, max-win and max-profit caps, and behaviour when a win would exceed the cap.
Test the failure paths that cost money connection dropped at bet submit, bet placed but the response lost, cash-out fired at the instant of a crash and confirm the balance reconciles afterwards.
Verify the bet history and the fairness-verification page tell the same story as the API and the chain.
Test streamed dealer tables end to end, from lobby listing through to a settled round and the transaction that follows it.
Pin down bet-window timing: when the window opens and closes against the server clock, whether a last-instant bet is accepted or rejected cleanly, and what happens to a bet that lands after the cut-off.
Verify game state stays in sync with the studio: dealer action versus client state, the displayed result versus the provider's round result, and a round ID you can trace from the table to the wallet.
Cover the table games properly blackjack split, double, insurance and bet-behind; roulette inside, outside, neighbour and racetrack bets; side bets such as perfect pairs and 21+3 each checked against its own payout rules.
Test chips and controls: denominations per currency, rebet, double, undo, table minimums and maximums, and multi-seat or multi-table play.
Break the real-time layer deliberately: socket dropped mid-round, reconnect, mobile backgrounding, airplane mode, throttled bandwidth, and a frozen stream while the round continues underneath then confirm the balance and outcome are still correct.
Check the stream itself: latency against game state, quality tiers and auto-degradation, audio, fullscreen, orientation change and picture-in-picture, on real devices and in in-app wallet browsers.
Validate third-party provider integrations: lobby listings and live stats, launch token and session handling, currency and limit mapping, error codes, and returning to the lobby cleanly.
Confirm balance sync between our wallet and the provider session, including funds held in a pending round and what happens if the session expires while a bet is live.
Verify round history and video replay match the settled result, and that responsible-gaming limits and jurisdictional blocks apply correctly mid-session, not just at launch.
Own test strategy for the titles you're responsible for and be accountable for the release call.
Run functional, regression and exploratory testing across desktop, mobile and tablet, on real devices as well as standard browsers.
Check the things regulators and players both care about: responsible gaming controls, bet limits, jurisdictional restrictions and correct behaviour across currencies and locales.
Write and maintain automated tests in our Playwright/TypeScript suite page objects, settlement specs and regression coverage for each new title and raise the bar on how we approach automation as a team.
Drive our AI-assisted and agent-led testing: use AI agents to accelerate test authoring, extend exploratory coverage, maintain suites against UI churn and triage failures then turn what works into a repeatable practice the rest of the team can use.
Judge where AI helps and where it doesn't. Agent-generated coverage still has to be reviewed, and a confident wrong answer about a payout is worse than no answer.
Support UAT and monitor titles after release verify fixes in production, and help triage live incidents when something behaves unexpectedly with real money on the table.
Raise clear, reproducible bugs in Jira and drive them to resolution with product owners and developers.
Investigate mobile stability issues crashes, memory pressure, rendering failures and get to a root cause rather than a workaround.
Spot patterns across the portfolio. A defect that appears in four games usually isn't four bugs.
Mentor and set standards. As the QA function grows, how we test is partly your design.
Communicate release readiness honestly: what's tested, what's risky, what isn't ready.
5+ years in QA on web or mobile products, with real ownership of test strategy rather than just execution.
Hands-on experience testing at least one of our three core areas slots, crypto/originals games, or live casino and the appetite to own all three.
Comfortable testing real-time, event-driven products: websocket traffic, state sync, timing windows and recovery after a dropped connection.
Strong hands-on automation with Playwright, Cypress or Selenium, and genuine comfort in TypeScript or JavaScript. You don't need to be a developer, but you should be able to design, maintain and refactor a suite yourself and explain your choices this is the area where we want you to make us better.
Real enthusiasm for AI-assisted testing, and hands-on use of AI tooling in your own workflow writing specs, generating test data, triaging failures, reviewing coverage. We want someone who has already started, not someone who intends to.
An appetite for learning that shows in your history. This space is moving faster than any of us can keep up with; we want the person who reads the changelog, tries the new thing on a real problem, and tells the team what actually worked.
The instinct to verify against the source of truth. If the balance on screen says one thing and the API response says another, you already know which one you trust.
An understanding of why floating-point arithmetic and money don't mix.
Strong exploratory testing the ability to break something that passed every scripted test.
Experience testing across multiple browsers, devices and screen sizes.
Confident with Jira and TestRail (or equivalents), and able to write a bug report a developer can act on without a follow-up conversation.
Able to work independently in a small team, comfortable saying when something isn't ready to ship, and able to hold that line with a product owner.
Experience running AI agents as part of testing agents that assist you, or agents you've handed a meaningful part of the job to. Agent-driven exploratory runs, self-maintaining or self-healing suites, LLM-generated specs reviewed into the repo, MCP-connected tooling, agentic failure triage. Tell us what you built, what it caught, and where it let you down.
A point of view on the limits: how you validate agent output, how you stop a suite from rotting, what you'd never let an agent decide on its own.
Live casino or live dealer experience, and integration testing against third-party game providers (Evolution, Pragmatic Play Live, Ezugi or similar) lobby, launch, limits, history and replay.
Video streaming QA: latency, degradation, reconnection behaviour.
Experience with real-money, payments, sportsbook, trading or other settlement-critical systems.
Crypto or blockchain exposure wallets, on-chain transactions, token precision.
Familiarity with provably-fair mechanics (server/client seeds, nonces, hash verification).
Casino domain knowledge: RTP, paytables, volatility, bonus mechanics, responsible gaming controls.
Social casino or sweepstakes-model experience.
Cloud device testing (BrowserStack or similar).
CI/CD experience getting a suite to run reliably on every build rather than on demand.
ISTQB / ISEB Foundation or equivalent.
