SENIOR SOLUTION / DATA ARCHITECT – FRAUD & FORENSICS
Level: Senior / Lead Architect
CapCircle is recruiting a Senior Solution / Data Architect for a fintech fraud detection and model modernisation programme.
This role requires a hands-on architect with proven experience designing production data and machine learning platforms in a regulated financial environment. The successful candidate will lead the architecture workstream and own the target-state design for a scalable, multi-model fraud detection platform.
ROLE OBJECTIVE
Define and guide the implementation of an industry-standard architecture that can host multiple fraud models, data stores and supporting services on shared infrastructure. The architecture must support recurring and low-latency detection, isolate production scoring from retraining workloads and provide a reusable blueprint for future scaling and cloud migration.
KEY RESPONSIBILITIES
- Lead the architecture assessment and define the target-state solution and data architecture.
- Design a multi-model-per-server architecture supporting ingestion, feature engineering, graph processing, model scoring, inline decisioning, model selection and case management.
- Define a two-instance PostgreSQL topology that separates detection and investigation workloads from offline feature engineering, graph computation and retraining.
- Design the authoritative transaction store, case-management store and dedicated offline feature store.
- Define an online feature store and low-latency inline decisioning service using Redis or equivalent technology.
- Ensure offline and online feature definitions remain consistent and support train-and-serve parity.
- Define data partitioning, retention, indexing, backup, access-control, audit and reconstructibility requirements.
- Establish policies for incremental, idempotent data ingestion and the handling of late-arriving records.
- Design CPU, memory and I/O isolation policies to prevent resource contention between detection, feature backfills and model retraining.
- Define orchestration, scheduling, sequencing, retry logic, observability and failure-isolation requirements.
- Establish latency budgets, availability targets, failover behaviour, default actions and graceful degradation for inline decisioning.
- Define volumetry and capacity requirements for data stores, feature processing and model-serving workloads.
- Ensure the architecture supports point-in-time correctness, model versioning, data lineage, evidence retention and regulatory audit requirements.
- Guide concurrency, load, sustained-cycle, peak-volume, backlog-recovery, replay and failover testing.
- Produce a scalability roadmap and cloud migration approach.
- Ensure the design can be extended to additional fraud typologies, models and business environments without reworking the core architecture.
- Work closely with the Graph Data Scientist / ML Lead, MLOps Engineer, Data Scientist and fraud investigation stakeholders.
- Provide clear architecture documentation, implementation guidance, technical governance and formal handover.
ESSENTIAL EXPERIENCE AND TECHNICAL REQUIREMENTS
- Proven experience architecting production-grade data, analytics or machine learning platforms within banking, fintech, payments or regulated financial services.
- Strong solution and data architecture experience covering batch, streaming and low-latency processing.
- Advanced understanding of PostgreSQL architecture, schemas, partitioning, indexing, workload separation, retention and performance optimisation.
- Experience designing offline and online feature stores for production machine learning.
- Experience with multi-model infrastructure, containerised workloads and orchestration frameworks.
- Strong understanding of Docker, Airflow or equivalent technologies.
- Knowledge of Redis or equivalent low-latency stores and Kafka or equivalent streaming platforms.
- Experience designing real-time or near-real-time decisioning solutions with clear latency and availability requirements.
- Understanding of graph data processing and the infrastructure required for graph feature computation.
- Strong knowledge of security, privacy, access control, auditability, data lineage and governance in regulated environments.
- Experience defining cloud migration, scalability, resilience and disaster-recovery roadmaps.
- Strong knowledge of fintech fraud typologies and MLOps best practices.
- Ability to translate complex business, fraud and technical requirements into practical architecture decisions.
ADVANTAGEOUS EXPERIENCE
- Architecture experience on fraud detection, financial crime or transaction-monitoring platforms.
- Experience designing investigator case-management and model-feedback integrations.
- Experience developing repeatable platform blueprints across multiple markets or business units.
- Knowledge of model-risk governance, drift monitoring and controlled model promotion.
WAYS OF WORKING
- Work within an Agile technology team focused on continuous delivery.
- Participate in sprint planning, daily stand-ups, sprint reviews, retrospectives and demonstrations.
- Own architecture decisions, deliverables, risks and dependencies throughout the engagement.
- Work collaboratively with technical teams while providing firm architecture direction and governance.
WHO SHOULD APPLY?
- This role is suited to a senior architect who has personally designed complex production data or ML platforms in a regulated financial environment. Candidates must be able to demonstrate hands-on architecture ownership beyond high-level conceptual or enterprise architecture work.
- Apply through CapCircle. Your CV must clearly show the platforms you have architected, the technologies used, the scale and regulatory environment, and your direct contribution to production delivery.
ADDITIONAL MANDATORY REQUIREMENTS
- A minimum of 5–8 years’ relevant professional experience is required.
- Candidates must hold a relevant tertiary degree and applicable professional or technical certifications related to the position.
- These are senior and advanced-level positions and are not suitable for interns, recent graduates or students.
- Candidates must be able to work on a hybrid basis from the office in Roodepoort, Gauteng, South Africa.
- Office attendance is required three days per week and is non-negotiable.
- Please only apply if you meet the required experience, qualification and hybrid working requirements.