Lead Risk Architect - Financial Crime Platform

Lead Risk Architect – Financial Crime Platform

TymeX, Johannesburg, Gauteng, South Africa

TymeX is the Technology & Product Development Hub of Tyme Group, with a global mission to become serial bank builders. Our Financial Crime platform plays a critical role in protecting customers and the bank from fraud, money laundering, and other financial crimes using advanced data, analytics, AI, and automation.

We’re looking for a Lead Risk Technology Architect with a Non-Financial Risk background to join our Financial Crime team. You’ll be part of the group responsible for designing, building, and maintaining high-performance AML & fraud detection systems, entity resolution, and compliance analytics across Tyme’s ecosystem.

The goal of the Financial Crime Platform is to reduce friction, improve outcomes for genuine customers, reduce costs and reduce vendor dependency, by architecting a unified non-siloed platform across Fraud and Financial Crime for the whole Tyme Group entities.

You’ll work with large, complex transactions and customer behaviour, applying models and decisioning frameworks in Databricks, integrating data from multiple systems in real time and batch, and helping design the data foundation for real-time risk detection and case management.

• Demonstrated experience in multiple forms of data ingestion, event-based streaming, synchronous real time ingestion approaches
• Architect our SLM (localized LLM) strategy
• Use the above as the basis for building a low cost, but high performance multi-agent AI architecture with the ability to create an ontology via knowledge graphs
• Design the ODS LDM and the Star/Snowflake LDM to support knowledge storing, sharing, audit and re-use for Agentic AI to store, forward, capture knowledge and decisions made.
• Apply entity resolution techniques in place, without further data movement
• Be a proponent of not moving large scale data movement to multiple data sinks, process data in place rather than move data to different sinks
• Be a proponent of the SHIFT LEFT paradigm and be a Data Foundation proponent
• Utilize tooling that allows SHIFT LEFT, for example instead of a physical Graph Database, a virtual Graph Database sitting on top of a solid Data Foundation.
• Understand real time scaling and performance needs for fast track processing, near real time and slow track batch oriented needs and which types of Compute, GPU and storage to apply and appropriate Notebook technologies (Scala, PySpark, Python, SQL)
• Understand how to construct and deploy natively or via third party LLMOps and MLOps and how a comprehensive Data & AI monitoring strategy can be put into place
• Be a proponent of Governance, Lineage, Auditability, Data As A Product (Data Contracts) via utilizing the power of the Unity Catalog across the entire architecture
• Design systems to lower cost, require less IT support and to be adaptable/flexible.
• Collaborate closely with Data Scientists, Analysts, Compliance, Operations and our Product/Feature teams to operationalize models and rules
• Integrate with external systems (e.g. sanctions, &V, biometrics and authentication systems) to enrich risk and identity data by building a rich profiling layer capable of supporting detection and investigation. Profiling for sub-millisecond needs in fraud detection architecturally and for slower profiling for investigations

• Bachelor's degree or above - ideally in Computer Science with specialization in Data Architecture/AI
• 10-15 years of experience, preference in Financial Services, Fast Moving E-Commerce industries
• Held the position of Lead Risk Architect at one of the above industries with a suitable size customer base and designed global/multi-country systems
• Hands on experience demonstrating architecture, design and implementation of Financial Crime or Cyber Security detection systems using Databricks, Snowflake, GCP Big Query/Table and/or Cloudera type systems. Systems designed on other architectural or older architectural paradigms are not what we are looking for, but a migration of knowledge from these to modern systems is critical
• Experience working with structured and…