Selected projects across financial services, higher education, and regulated SaaS. Each case study covers the problem, the architectural approach, security and compliance considerations, and what I'd do differently today.
Modernization · Financial services
Migrated a three-decade-old bash-script batch system to event-driven serverless on AWS. Owned the engagement end-to-end: target-state design, proof-of-concept, CI/CD pipeline, infrastructure, data lifecycle and retention design, DPIA and financial GDPR compliance, UAT, productionization, and decommissioning of the legacy stack. Mentored two junior engineers through the build.
Full case study coming soon.
Data platform · Enterprise
Designed and delivered an enterprise data warehouse consolidating SAP and operational systems into Redshift, with Glue-based ETL and S3 as the data lake. Included dimensional modelling, performance tuning, and cost optimization.
Full case study coming soon.
Security & compliance · Payments
Authored payment integration documentation and PCI-DSS-aligned architecture for a regulated payments flow. Tokenization, scope reduction, KMS-based encryption, segmentation, and audit-trail design.
Full case study coming soon.
AI & governance
Authored internal AI data governance policy covering data sourcing, model usage boundaries, privacy, IP, and operational controls for generative AI adoption in a regulated environment.
Full case study coming soon.
AI engineering · Proof of concept
Designed and built a retrieval-augmented generation system using AWS Bedrock and Titan embeddings, with metadata-driven retrieval over relational tables — letting users query operational data in natural language without exposing schemas to the LLM.
Architecture has evolved since — write-up coming once the current version is shipped.