Library of Alexandria

A private research corpus for studying mythology, folklore, religion, and cross-cultural primary texts. The real work is provenance and curation, not accumulating the largest folder.

privatenot publicly distributed
curatedquality over volume
sourcedprovenance retained
localresearch infrastructure

The question

Can a local research system reason from primary and carefully edited sources instead of shallow summaries? The project began as a collection problem and became a data-quality problem.

Curation before scale

Early versions contained duplicates, mislabeled material, weak metadata, and sources that did not belong. The important milestone was not making the corpus larger. It was removing untrustworthy material and recording enough provenance to explain what remained.

Responsible sourcing

Collection respects source terms and explicit opt-outs. Material is checked for identity and provenance before it becomes part of the research set. A download succeeding is not evidence that the content is correct or appropriate to retain.

What it supports

The corpus supports private retrieval and model experiments where answers need to remain connected to source material. The public proof is the curation method and the lessons learned. The corpus contents, collection manifests, storage layout, and training recipes remain private.

The durable lesson is simple: data quality work is mostly subtraction, provenance, and the willingness to reject impressive-looking volume.

// private
The corpus is not publicly distributed. For research or hiring conversations, I can discuss the curation and verification process without exposing the dataset or collection system. Get in touch.