Intelligence

An assistant that works from your archive

Nolocron's assistant isn't trying to out-think frontier models. It does what they can't: work from your governed archive, under the scope you set, and show its work.

AvailableAssistant

A cited answer, and what's underneath

The answer

Pricing moves to three tiers in Q3 S1, with the team aligned on a usage-based add-on S2.

S1Meeting notes · Mar 3
S2Email from Dana · Mar 5

Each handle is validated and linked to the real block.

What's underneath

Derived — generated, labeled, removable

SummaryAuto-tagsEmbedding

Canonical

Your original record

Never overwritten. Always recoverable underneath.

Follow every claim to its source — and the model's best guess never becomes your record.
01

A general assistant is brilliant but starts every session from zero — it can't reach your prior work, prove what it used, or build on yesterday. Memory without governance is a liability; governance without memory is useless.

02
  • Streaming chat with multi-turn memory, working from the slice you point it at.
  • Every tool call runs through a governed action layer, with human approval for sensitive operations.
  • It cites the blocks it drew from, linked back to the source.
  • Every run is recorded, so you can see exactly what it touched.
03
  1. Choose the source or project the assistant may see.
  2. Ask a question that needs your prior knowledge.
  3. Read the answer with its citations.
  4. Open the run record to see what it drew on.
04
  • Governed tool use, not open-ended database access.
  • Citations and receipts make its work reviewable, not magic.
  • It's one way to reach your archive — not the reason the archive exists.