System design
Architecture breakdowns that make estimation, failure modes, and engineering choices legible.
Learning paths · full references, honest progress
The map is complete.
My progress isn't.
This is not a finished curriculum. It is a live record of what I am learning now, what comes next, and what is still only an intention.
In progress
2 paths in motion
Daily problem-solving and AI inference engineering are active. New material appears only after it has been studied, tested, and explained.
updated as I learnAll roadmaps
Reference maps and my route through them
2 active · 1 next · 4 later
Daily problem solving
Algorithms, Java internals, backend architecture, and low-latency systems practiced as one continuous engineering discipline.
- Algorithmic foundations
- Java foundations
- Advanced structures
- Spring and enterprise Java
AI systems mastery
The complete path from ML foundations and transformer internals to efficient, observable, distributed AI infrastructure.
- Machine-learning foundations
- Transformer internals
- Retrieval systems
- LLM applications and agents
Production AI + data
Applied RAG, agents, evaluation, Spark, Delta Lake, MLflow, and data systems assembled into production-grade architectures.
- RAG architecture
- Spark and Delta Lake
- Stateful agents
- Evaluation platform
Deployed engineering
The discipline of discovering ambiguous problems, owning integrations, delivering under constraints, and learning from real usage.
- Operating foundations
- Domain immersion
- Discovery and scoping
- Integration delivery
Staff systems
Large-scale architecture, reliability, performance, technical leadership, and communication developed through artifacts and operating practice.
- Large-scale system design
- Architecture thinking
- Reliability engineering
- Performance engineering
Business systems
Strategy, finance, marketing, operations, leadership, economics, entrepreneurship, and negotiation learned through real application.
- Strategy
- Finance and accounting
- Marketing
- Operations
Publishing surfaces
Forward-deployed engineering
Generalised field notes from customer-facing engineering—never employer, client, or confidential detail.
Learning labs
The evidence layer: experiments and mistakes that show exactly what changed the mental model.
Knowledge archive
Durable outputs earned through the journey—not a library filled before the learning happens.
Publishing protocol
Learn → build → explain.
No card becomes a lesson because it was added to a list. It moves only when there is evidence: a question, an experiment, a diagram, a benchmark, a mistake, or a clearer explanation.
- 01Question
- 02Study
- 03Visualise
- 04Test
- 05Explain
- 06Revisit
Visual lab · planned capability
Concepts should move,
not sit on a page.
As the journey progresses, difficult ideas will become animated diagrams, mathematical explainers, and small simulations—built to make systems intuition visible, in the spirit of visual-first teaching.
request
system
insight
Interactive visualisation · coming with the concepts