Selected Works

A curated list of implementations, experiments, and systems.

A custom-built live monitoring and autoscaling daemon for the KNUST Cloud, ensuring high availability across all cluster nodes.

Go-based daemon that monitors Swarm service load and adjusts replica counts in real time based on Prometheus metrics.

High-performance reverse proxy configuration and zero-trust identity layer for distributed microservices running on Docker Swarm.

The original hand-rolled infrastructure monitoring system, now decommissioned and replaced by Production Engine.

A reactive management dashboard for OpenStack clusters built with Nuxt 3 and TypeScript, providing real-time visibility into cluster state.

Low-latency data replication across edge nodes using WebSockets and dedicated backplanes, targeting sub-50ms propagation.

Real-time telemetry pipeline for high-throughput sensor data using Go and Redis Streams, processing over 50k events per second.

Distributed search engine layer built on top of Elasticsearch, providing sub-100ms full-text search across multi-tenant document stores.

Zero-trust identity layer with JWT-based authentication, role-scoped API tokens, and biometric second factor for administrative consoles.