- Location
- Cairo, Cairo Governorate, Egypt · published
- Department
- IT
- Source
- Workable
Description
About the role
You will be the analyst embedded in one of our product squads — the person who turns product questions into numbers the company can act on. Every meaningful product change at invygo ships behind an experiment, and you own the experiment: sizing it before launch, catching what's broken while it runs, and delivering the verdict when it ends. Your verdicts gate ship/kill decisions, so independence and statistical honesty matter more here than in a typical analyst role.
You'll work in a heavily AI-augmented setup: our analytics runs through an AI agent the whole company uses to query governed data, and analysts here operate and improve that agent. AI does the routine work; you do the judgment.
What you will be doing
Experimentation — the core of the role
● Own A/B tests end to end: design and sizing (minimum detectable effect, power, choice of unit and metric), monitoring while the test runs (assignment sanity checks, rollout dilution, contamination between arms), and the final effect estimate with honest uncertainty.
● Handle the messy reality of real-world experiments: small samples, ratio metrics, outliers that can flip a verdict, shared car inventory letting arms interfere, and value metrics that mature over weeks.
● Apply variance-reduction techniques (e.g. CUPED, stratification) where they genuinely help — and recognize when reaching for them after the fact is method-shopping.
● Deliver verdicts that hold up: no p-hacking, no metric switching after the fact, clear significant / not significant / can't-know-yet calls.
Product analytics
● Be the data voice inside your squad: define metrics for new features, instrument them correctly with engineering, and keep product changes reflected in the analytics layer and core reports.
● Investigate funnels, cohorts, retention and pricing questions across UAE, KSA and Qatar — from raw event data to a recommendation.
AI-augmented analytics
● Use and improve our internal AI analytics agent: extend its context and documentation as the product evolves, review its outputs for correctness, and grow the evaluation set we score it against.