Antifraud Scientist at Higgsfield AI

Higgsfield AI, a rapidly growing AI company with $1B ARR, seeks an Antifraud Scientist to develop anti-fraud models.

Develop an anti-fraud model. Context: Over 6 million generations per day; each fraudulent generation consumes real GPU, so the model must stop abuse before computational resources are spent. The adversary adapts within days: farms with hundreds of accounts, resale of access, automated collection of promo bonuses. The system must withstand this race.

At least 3 years of experience in DS/ML, Python and SQL, with models in production. Experience with boosting, class imbalance, and link graphs. Ability to evaluate the model through prevented losses and cost of false blocks. Willingness to analyze cases manually and collect signals from raw events, payments, and user behavior. Self-sufficiency: from initial hypothesis to working scoring, monitoring, and model updates. Willingness to work 6 days a week.

Office in Almaty. Full relocation support (financial and organizational). Salary $100–350K per year.