ARCHIVE_VOL.2 / ENGINEERING
EngineeringProjects
Projects built to explore new technologies, solve meaningful problems, and deepen my engineering expertise.
Where Production Systems reflects what I've delivered professionally, this is what I build when I'm choosing the problem myself — driven by curiosity and ownership rather than a client requirement.
Dyslexia Management Platform
Overview
A platform designed to support learners with dyslexia through adaptive, accessible learning experiences — built to make reading intervention progress visible to both learners and the educators and caregivers supporting them.
Problem
Reading support tools are rarely built accessibility-first. Most retrofit accommodations onto interfaces designed for typical readers, and progress tracking is often scattered across paper records or disconnected spreadsheets, leaving educators without a clear picture of how a learner is actually progressing.
Architecture
Client
React interface with accessibility controls built into the rendering layer, not layered on top.
API Layer
Serves structured content and handles progress data for each learner.
Data Store
Persists learner progress, content structure, and session history.
Engineering Challenges
Accessible By Default
Font, spacing, and contrast needed to be first-class settings, not an afterthought accessibility panel.
Adaptive Content
Structuring reading material so it could be presented differently per learner without duplicating content.
Progress Across Sessions
Designing a data model that could track intervention progress meaningfully over time, not just per session.
Room For Personalization
Building the architecture so a future recommendation layer could plug in without a rewrite.
Key Features
- Adjustable reading settings — font, spacing, and contrast
- Progress tracking across sessions
- Structured, adaptable intervention content
- A dedicated view for educators and caregivers
Tech Stack
Lessons Learned
This project pushed me to think about accessibility as an architectural decision, not a styling pass at the end. Designing the content model to support different presentations per learner — before I knew exactly what those presentations would be — was the hardest and most useful part of the build.
Future Roadmap
Bundesliga Analytics Platform
A data engineering project exploring large-scale football analytics — pulling together Bundesliga match and spending data through a distributed processing pipeline to explore questions like whether spending actually predicts results.
The interesting engineering problem wasn't the charts at the end — it was the pipeline getting there: ingesting and cleaning inconsistent match data, structuring it for querying at scale, and deciding what to process in batch versus what needed to stay queryable on demand.
Engineering Focus
Technologies
In Progress
Currently Building
Projects I'm actively working on to deepen my expertise in AI, robotics, and intelligent software systems. Check back — these will keep changing.
Gesture-Controlled Drone
An intelligent drone controlled through hand gestures, using computer vision and machine learning to translate movement into flight commands in real time.
Currently — Prototype
Current Milestone
Building the real-time gesture recognition pipeline and testing it against onboard flight controls.
What's Next
Expanding the gesture vocabulary, tightening latency for real-time response, and optimizing inference to run fully on embedded hardware.
Focus Areas
hushHush Recruiter
An AI-driven recruitment platform exploring how to automate candidate discovery, evaluation, and hiring workflows — analysing publicly available developer signals, notifying shortlisted candidates, and issuing time-limited coding assessments.
Currently — Prototype
Current Milestone
Building the candidate evaluation flow, including time-limited coding assessments and a review interface for hiring managers.
What's Next
Refining privacy-conscious handling of candidate data, and building out the hiring-manager review dashboard.
Focus Areas
Recurring Threads
Engineering Themes
Areas I keep finding myself drawn back into — worth reading as exploration, not a claim of mastery.
Right Now
Currently Exploring
LLMs
Prompting, evals, and application design
RAG
Retrieval-augmented generation pipelines
AI-Assisted Development
Using AI tools as part of the engineering workflow
Cloud Deployment
AWS fundamentals, containerized deploys
MLOps
Getting models from notebook to production
Agent Workflows
Tool-using, multi-step AI systems
END_OF_LOG