Projects

Things I've built and am building.

Anima Mundi // Obsidian PKM
Active

Anima Mundi // Obsidian PKM

Self-organising Personal Knowledge Management System. Semantic search across clipped articles, videos, PDFs, GitHub repos, and links. Private repo — going public soon.

PythonChromaDBNLPembeddingsfastembed
Recall // Document Q&A
Planning

Recall // Document Q&A

Queryable knowledge base over personal document collections. Natural language Q&A across PDFs, notes, and clipped sources via RAG and vector retrieval.

PythonChromaDBRAGNLPembeddings
Lexis // Text Classifier
In progress

Lexis // Text Classifier

Fine-tuned transformer pipeline for multi-label document classification. Built for low-resource domain adaptation.

HuggingFacespaCyPyTorchNLP
Strata // Data Pipeline
In progress

Strata // Data Pipeline

Modular ELT pipeline with schema inference, lineage tracking, and incremental loading. Designed for reproducibility.

PythonDockerPostgreSQLdbt
Meridian // Embedding Explorer
Planning

Meridian // Embedding Explorer

Interactive visualisation of high-dimensional embedding spaces. Dimensionality reduction with UMAP and clustering overlays.

PythonUMAPChromaDBfastembed
Aqua // Marine Bioremediation
Planning

Aqua // Marine Bioremediation

Queryable sensor data infrastructure for marine restoration and bioremediation monitoring. Ingesting environmental sensor streams into a searchable time-series database for reef and coastal ecosystem tracking.

PythonPostgreSQLIoTtime-seriesRAG