How we compare

Structured beats plausible.

A chatbot writes a fluent answer. GRADE enforces the spec, then writes the answer — and shows the line it came from. This is a category comparison, not a callout of any one tool.

The categories

GRADE Chatbot Vector search Spreadsheet
Enforces a hard spec exactly
Same screen result every time
Explains every elimination
Quotes your own data sheet
Won't invent a product
Keeps the human as approver

We tested the chatbots

We didn't assert this — we measured it, against the strongest chatbot on the same raw data sheets a rep works from.

How we test our own engine →
RECALL
17%
of the correct grades surfaced from raw data sheets.
STABILITY
7 of 12
inquiries got a different answer when asked again — even from a clean table.
BY DESIGN
Survivors only
the model never sees a product that failed the screen, so a reply can only name grades from your catalogue that passed.
RULE
Reads
the model reads; code decides; the human approves.

Why the difference is structural

The categories above differ in where the decision lives. In a chatbot, the model decides. In vector search, a similarity score decides. In GRADE, the model never decides at all.

01
A model reads
It turns the inquiry's prose into typed constraints — and does nothing else on the screening path.
02
Code decides
Every hard constraint is a SQL filter over typed columns. The same spec against the same catalogue gives the same verdicts, with a reason on every elimination.
03
A human approves
The rep sees the constraints, the survivors and the receipts, edits the draft, and sends it — or doesn't. Nothing leaves on its own.

The honest section

When not to use GRADE

If any of these is you, GRADE is the wrong tool, and we would rather say so now.

  • Your specifications are not published in data sheets. If the numbers only exist in someone's head or an internal system, we cannot screen on them.
  • You need field service, formulation development, or lab work. We answer inbound inquiries. We do not visit plants.
  • Your inquiries are relationship-led rather than specification-led. If customers call and ask "what should I use," with no stated requirements, there is nothing to screen against.
  • A general assistant already handles your catalogue well enough. Smaller catalogues sometimes do not need this.

This is not about team size. A small team gets real value from GRADE too.

See the difference

Screen the same inquiry both ways.

Paste one on the home page and watch GRADE enforce the spec — then ask a chatbot the same thing and compare.

Try the live demo → Talk to us