ShAIne
Corpus Growth
The RAG corpus powering Interview ShAIne — content files, chunk counts, and how the knowledge base has grown over time.
Total chunks
306
Content files
30
Last ingest
23 days ago
May 1, 2026
Chunks over time
project milestonemethodology change
Milestones
Annotated entries from content/corpus-history.json. Automated snapshots are recorded by the ingest script.
| Date | Chunks | Files | Chars (K) | Chunk sz | Note |
|---|---|---|---|---|---|
| March 18, 2026Site launch | 56 | 5 | 33.6 | 600 | Initial scaffold: bio, resume, skills, about, 1 reference |
| March 18, 2026 | 79 | 8 | 47.6 | 600 | Full site launch: job-preferences v1, transitions, hdtvmagazine |
| March 18, 2026 | 96 | 9 | 57.7 | 600 | Added interview-voice.md — voice training begins |
| March 23, 2026 | 131 | 16 | 78.5 | 600 | Scaffolded mission-and-motivation, 7 new files |
| March 24, 2026job-preferences expansion | 172 | 17 | 103.4 | 600 | job-preferences.md: 3K → 32K chars (+25% of total corpus) |
| March 24, 2026 | 203 | 18 | 121.6 | 600 | Added notes-domino-background.md, certifications |
| March 25, 2026 | 209 | 19 | 125.7 | 600 | fit-config.json added to context; first blog post published |
| March 29, 2026 | 244 | 20 | 138.5 | 600 | Published: Under the Hood — ShAIne RAG pipeline |
| April 15, 2026 | 281 | 21 | 153.3 | 600 | Published: Leading in 4K |
| April 17, 2026 | 299 | 22 | 161.1 | 600 | Published: Ground-Level HDTV Magazine Revival (15-week build log) |
| April 21, 2026 | 285 | 25 | — | 1000 | Paragraph-aware chunking introduced; chunk size raised to 1000; Bellese content updates |