Data and memory
Where the context lives: knowledge bases, search, agent memory, working with documents. materials: 6
Knowledge base
Astro: a static site that can handle thousands of pages
Astro assembles each page into ready-made HTML at the build stage and does not load JavaScript into the browser by default. Analysis of mechanics: routes from files, collections with a schema, islands, assembly in dist/ - and where this approach stops working.
Notion as a data source for the site
The Notion database sends page properties and a tree of blocks via the API, the site takes them in the assembly and turns them into static pages. Analysis of the pipeline, pitfalls with expiring links to images and request limits.
Obsidian as a knowledge base with which the agent works
The Obsidian repository is a regular folder with markdown files, so it can be read equally by a human, a script, and an AI agent. What to put in frontmatter, how to give the agent access, how to prevent him from littering the database with duplicates, and how to check that the knowledge is there at all.
Answers to the same type of customer questions without an operator
How it works: classify requests by intent, build a knowledge base from operators’ answers, and let the bot reply only with a retrieved passage and source link. It hands everything else to a person with the conversation history.
Vector search: why it is needed and when it is not needed
Embeddings, document slicing, ANN indexes and a hybrid with full-text search - the whole mechanics. Separately: a list of tasks where a vector is inferior to a regular word search, and how to measure this on your corpus.