Senior Data Scientist Transaction Intelligence

Malaa Technologies; Riyadh; On-site; SAR 260,000 - 420,000; Full time

Job summary

Malaa Technologies in Riyadh is seeking a data scientist to transform raw transaction data into actionable insights, spanning enrichment, categorization, and risk modeling.

You will ship production-grade models, collaborate with product and engineering, and navigate fintech governance, privacy, and compliance while handling bilingual text and large-scale pipelines.

Qualifications

• 3+ years of applied ML/data science with production experience.
• Strong analytical skills: EDA, statistics, feature design, SQL.
• Hands-on with text similarity, fuzzy matching, or entity resolution on noisy strings.
• Proficiency in Python and scientific stack with performance awareness.

Responsibilities

• Enrich and categorize transaction data, resolving merchant strings to real entities.
• Own end-to-end enrichment models: design, build, evaluate, and maintain production code.
• Partner with product and engineering to design, ship, and measure model-backed features.
• Level up enrichment foundations for unseen merchants and tolerant name matching.
• Manage model confidence and calibrated probabilities with principled abstention.

Skills

SQL

Python

Applied ML

Text processing

Entity resolution

Data governance

Collaboration

Tools

scikit-learn

SciPy

NumPy

Job description

About the Role

Malaa is Saudi Arabia’s first retail Open Banking platform. The data scientist will transform raw transaction behavioral data into actionable insights, spanning the full pipeline from enriching transactions and categorizing noisy merchant strings in mixed English and Latinized Arabic to building models that drive risk assessment, anomaly detection, and pattern recognition across our services.

Responsibilities

• Enrich and categorize transaction data, resolving merchant strings to real merchant entities.
• Own end-to-end enrichment models: design, build, evaluate, and maintain production code.
• Partner with product and engineering to design, ship, and measure model‑backed features.
• Level up transaction enrichment foundations, ensuring generalization to unseen merchants and tolerant name matching.
• Manage model confidence, delivering calibrated probabilities, principled abstention, and confidence‑based routing.
• Build insight models for pattern recognition, anomaly detection, and behavioral analysis on transaction streams.
• Validate risk models through backtesting, monitor discrimination and calibration, and drive iterative improvements.
• Handle bilingual, informally Romanized Arabic text; manage truncation artifacts, bank quirks, and heavy‑tailed distributions.
• Build efficient, scalable batch pipelines at hundreds‑of‑millions‑record scale, re‑runnable as models improve.
• Design within regulated fintech guardrails, respecting data governance, privacy, and cybersecurity constraints.
• Develop and maintain evaluation discipline: labeled datasets, regression test suites, and metrics that answer whether a change helped.

Qualifications

• 3+ years of applied ML/data science with models shipped and maintained in production at scale.
• Strong analytical skills: exploratory analysis, statistical rigor, feature design, and SQL fluency.
• 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 performance awareness at scale.
• Demonstrated maturity handling data governance, privacy, cybersecurity, and compliance constraints.
• Track record of turning ambiguous complaints into measured, regression‑tested properties.
• Comfort reading and debugging model code produced by others and working directly with product teams on loosely‑defined problems.
• Nice‑to‑haves: Arabic/Arabizi text processing, credit or behavioral risk modeling, discrimination and calibration measurement, backtesting against realized outcomes, text embeddings, approximate nearest‑neighbor retrieval, LLM‑assisted labeling, and ML lifecycle and batch‑serving tooling.