I didn't start in AI. I started in the engine room of enterprise IT (firewalls, Active Directory, server infrastructure), where systems either work in production or they don't. That grounding is the thing most AI engineers never get.
An MSc in AI took me into RAG systems, anomaly detection, and computer vision. The operator's instinct stayed: a model that can't deploy, monitor, and survive real data isn't finished. Today I build at that intersection and study the governance patterns that keep it sustainable.
AI Engineer with an MSc in Artificial Intelligence and a track record of architecting end-to-end generative and agentic AI systems that ship to production. I work across LLMOps, RAG pipeline design, vector databases, and multi-agent orchestration with LangChain, LangGraph, and AutoGen, with a strong grounding in responsible AI: prompt-injection defence, PII handling, content moderation, and auditability.
- ·Designed and deployed ML classification and predictive models for institutional use cases, applying responsible AI principles (data governance, PII handling, auditability) across all models.
- ·Engineered Python workflow automations for asset and data management, reducing manual intervention by 40%.
- ·Administered enterprise IT infrastructure across hardware, software, and network systems, ensuring 99%+ availability for 500+ academic users.
- ·Led bulk user provisioning and lifecycle management within LMS platforms, standardising environments for security and performance.
- ·Authored and validated 5+ psychometrically sound certification exam items covering single-agent, multi-agent, and RAG-based AI systems (LangChain, LangGraph, AutoGen).
- ·Defined core competencies for agentic AI engineering including LLM orchestration and vector DB integration (pgvector, ChromaDB) through formal Job Task Analysis.
- ·Collaborated with psychometricians and SMEs to ensure technical accuracy and examination-standard coverage.
- ·Led end-to-end development of Oakbent's internal website framework with dynamic pages and CMS components.
- ·Built Python automation tools to streamline product data handling, reducing manual effort by ~30%.
- ·Maintained version control via GitHub; contributed to sprint planning in Notion and Trello.
- ·Engineered a custom ATS pipeline with Python and NLP resume parsing, processing 500+ applications/month and boosting selection efficiency by 60% while cutting time-to-hire by 25%.
- ·Automated multi-platform job distribution, increasing high-quality applicant volume by 80%.
- ·Built analytics-backed interview scoring models that improved hire quality by 65%, with Power BI dashboards for funnel monitoring.
- ·Managed IT infrastructure including Fortinet firewalls, Active Directory, and Windows Server, reducing unauthorised access incidents by 85%.
- ·Optimised PABX systems and inter-VLAN routing, cutting communication delays by 40%.