CxSTAT IQ · CxBrain

Ask the project anything.

CxBrain reasons over every equipment item, checklist, issue, test and document in the graph, and answers with its sources side by side. The project knowledge stops living in three heads.

app.cxstat.com/iq/cxbrain
The problem

The answers exist, the finding is the job.

Every question in the morning meeting is answerable from records the project already owns. It just takes forty minutes and the right person to find them.

person_search

Knowledge concentrates in people

Two or three people can answer anything, so everything routes through them. They become the bottleneck and the risk.

schedule

Questions cost meetings

What is blocking IST? Which valves changed in Rev C? Each answer is an afternoon of cross-walking logs, sheets and email.

psychology_alt

Chatbots guess

Generic assistants answer confidently without your registers, your history or any obligation to cite what they claim.

How it works

Answers with receipts.

CxBrain is grounded in the evidence graph. If it cannot cite it, it does not say it.

hub

01 · Grounded in the graph

Equipment, tests, issues, documents and their relationships, with lineage back to source pages.

psychology

02 · Reasons, not retrieves

Multi-hop questions get multi-hop answers: dependencies traced, histories compared, deltas explained.

format_quote

03 · Cites or declines

Every claim carries its sources. When the graph does not know, CxBrain says so instead of improvising.

What you get

The project brain, on demand.

For engineers, managers and owners alike.

forum

Plain-language questions

From “what is blocking IST?” to “which CRAHs share the header with the failed pump?” No query language, no tribal knowledge.

format_quote

Sources on every answer

Each response lists the registers, tests and documents behind it, one click from the exact page.

manage_search

Cross-record reasoning

Issues joined to tests joined to drawings joined to schedules. The connections humans forget to make.

shield

Project-scoped by design

CxBrain reads only your project graph. Answers never leak across projects or into model training.

Questions we get

Frequently asked.

How is this different from asking ChatGPT?add
CxBrain reads your live project graph, cites every claim, and declines when the record does not support an answer. A general chatbot has never seen your registers and will answer either way.
What can it actually see?add
Whatever your project graph holds: equipment, checklists, tests, issues, documents read by Vault, and their relationships, with project-scoped access control.
Can it be wrong?add
It can misread an ambiguous record, which is why every answer shows its sources. Verification is one click, not an act of faith.