About Me
Work Experience
▸Working in AI & Data at Accenture, designing and scaling modern data foundations — cloud platforms, lakehouse architectures, and the agentic pipelines that make them intelligent. Helping organisations move beyond passive data infrastructure to agent-driven ecosystems that turn strategy into measurable outcomes.
▸Working on agentic program intelligence for a large utility navigating complex transformation and regulatory compliance. A single AI-native intelligence layer across architecture, delivery, risk, and review, connecting JIRA signals, integration catalogues, solution artefacts, and compliance evidence into one grounded operating view.
▸Program Intelligence closes critical coordination and alignment gaps by replacing manual follow-up with scoped, persona-driven AI that surfaces blockers, gauges program health, and accelerates document authorship and review. Human-in-the-loop design preserves expert judgement for nuanced decisions, enabling faster course-correction, reduced analyst toil, and tighter alignment between strategy, execution, and compliance obligations.
▸Led the Unstructured Document Migration squad for the ANZ–Suncorp bank acquisition, collaborating with solution design and architecture teams to define parallel payload and metadata migration pipelines. Designed a Migration-Copilot agent to automate artefact generation — SDS, TDS, and source-to-target mappings — from ambiguous requirements. Advised ANZ SMEs on data lineage, migration flows, timelines, ADMP metadata transformations, and reconciliation design, while coaching analysts and establishing reusable templates to uplift delivery consistency.
▸At Westpac, designed Cloudera → Azure Synapse migration validation (HDFS→ADLS2→DSQL); built Orc→Parquet→DSQL lineage and CTM Recon tracking. Delivered transformations, semantic layer, and Power BI dashboard.
▸Delivered a Databricks–Power BI–MS Fabric asset analytics solution (30+ metrics) for an $8.5B portfolio enabling audit-ready reporting and faster month-end close.
▸Built Power BI–MS Fabric semantic models and an Investment Vetting data product for a $300B portfolio. Leveraged CoPilot Gen AI; streamlined deployment via GitHub CI/CD.
▸Defined requirements for a Workday data warehouse; designed Data Vault & dimensional models (Erwin); managed Azure Synapse for employee analytics.
▸Led BI and process automation in healthcare; created Power BI reports tracking funding, rostering, and client experience.
Projects
Single AI-native program intelligence layer connecting JIRA, architecture, risk, and compliance signals to surface blockers and align strategy with execution.
Interview-driven multi-agent system (17 Claude Sonnet subagents) generating fit-for-purpose architecture artefacts — ADRs, C4, threat models, NFRs, data models — via intent-signalled selection. Features bi-temporal memory consolidation, skill graduation gates, adversarial critique loops, and a quality-gated PlantUML→draw.io ECM diagram pipeline producing 18 enterprise-grade outputs.
Modular pluggable multi-model routing engine selecting the optimal LLM per task. Implements ReAct, Reflection, Deep Research, Tool-Use, planning, memory.
SWOT, metric design, BI strategy, predictive models. HD (93%) — Dean's List, 1st in cohort.
Researched synthetic data generation using GANs and VAEs; advanced AI techniques report (87%).
