Stocky AI
Decision Intelligence Agent
- Architected an Agentic AI + MCP system, dockerized to ECR/ECS with PyTest-backed CI/CD.

Each role and milestone is a stage in the pipeline — flowing top to bottom.
BSc (Hons) Computer Science — Data Analytics
Foundational training in distributed systems, databases, and statistical modelling — the upstream source feeding every project downstream.

Unified data platform · Spark on Kubernetes
Shipped to a platform handling millions of jobs. Worked across the Spark, Airflow, and Kubernetes stack with full CI/CD ownership.

The next stage of the pipeline is currently being provisioned.
Four infrastructure layers — each component is a tool I reach for to ship production systems.
Languages and runtimes I write production code in.
Building pipelines, orchestrating jobs, modelling state.
Shipping containers, infrastructure, and CI/CD.
Modelling, agents, and the visualisation layer on top.
Each project is a deployable service in the infrastructure grid — architecture, stack, and source code.
A few highlighted credentials — view the full set on the qualifications page.
A small cabinet — every trophy came with a real product behind it.
Drop a message, copy a channel, or talk to the terminal — whichever feels right.