About the role:
The company is Saudi Arabia’s first retail Open Banking platform in the Kingdom. Our product is built on one core asset: hundreds of millions of records of behavioral data. Turning that raw data into something actionable is the primary responsibility of this role, encompassing the entire pipeline. This includes enriching transactions (categorizing short, noisy merchant strings in mixed English and Latinized Arabic, and resolving them to real merchant entities), followed by building models and analyses that translate transaction data into insights across our services — such as categorization, pattern recognition, anomaly detection, and risk assessment.
Ready to apply for roles like this?
Unlock the company name and direct application link. Subscribers get instant access to fresh jobs across Dubai, Abu Dhabi and Riyadh, many with visa support.
Unlock employer & apply directly
As the data scientist for transaction data across the company, you will own the enrichment models end to end, handling modeling, evaluation, and the production code. You'll also partner with product and engineering teams to design, ship, and measure model-backed features in different services.
What you'll work on
- Own and enhance transaction enrichment, establishing a foundation for all subsequent work: creating categorization schemes that generalize across unseen merchants, and developing merchant name matching systems that can handle harmless variations without conflating genuinely different businesses.
- Ensure model confidence across our systems, including well-calibrated probabilities, principled abstention on uncertain cases, and confidence-based routing that products and review workflows depend on.
- Build transaction-based insight models for other services, focusing on pattern recognition, anomaly detection, and behavioral analysis of transaction streams.
- Own the ongoing validation of our risk models by backtesting scores against actual user behavior, monitoring discrimination and calibration as behaviors change, and driving model improvements based on data insights.
- Manage our text and behavior data as they exist: handling bilingual, informally romanized Arabic with unstable spelling; addressing transaction streams with truncation artifacts, bank quirks, and heavy-tailed distributions.
- Develop efficient batch pipelines that can manage hundreds of millions of records, designed for routine re-runs as models evolve.
- Create solutions within the constraints of a regulated fintech environment, adhering to data governance, privacy, and cybersecurity requirements. You will collaborate with teams responsible for these standards to ensure excellent solutions are developed without circumventing these regulations.
- Establish a robust evaluation discipline for these systems, including creating labeled datasets, regression test suites for model behavior, and metrics to assess the impact of changes on each release.
Must-have qualifications
- 3+ years of applied ML/data science experience, with models shipped and maintained in production at scale — including understanding their failure modes.
- A strong analytical range beyond modeling, encompassing exploratory analysis, statistical rigor, and feature design on behavioral/tabular data (SQL fluency assumed).
- Hands-on experience with text similarity, fuzzy matching, or entity resolution on noisy real-world strings.
- Proficiency in Python and the scientific stack (scikit-learn, scipy, numpy), with a focus on performance and cost efficiency at scale — including vectorization, sparse data structures, and efficient batch computation.
- Demonstrated ability to work within externally imposed constraints — including data governance, privacy, cybersecurity, compliance, and infrastructure policy — while collaborating with the relevant teams.
- A proven track record of addressing ambiguous complaints regarding model performance and turning them into measurable, regression-tested properties.
- Comfort in reading and debugging model code written by others, as well as collaborating directly with product teams on loosely-defined problems.
Nice-to-haves
- Arabic / Arabizi text processing experience is a strong plus; our data includes bilingual elements with unstable romanization.
- Knowledge of credit or behavioral risk modeling and validation, including scorecards, measurement of discrimination and calibration, and backtesting against realized outcomes.
- Familiarity with text embeddings and approximate nearest-neighbor retrieval in resource-conscious environments.
- Experience with LLM-assisted labeling, distillation, or weak-supervision pipelines.
- Proficiency with ML lifecycle and batch-serving tools (experiment tracking, data versioning, distributed task queues).
Employment Details
- Department: Engineering
- Employment Type: Full Time
- Location: Riyadh
- Application Deadline: 30 September 2026
