Measure answer quality, don't assume it
An assistant you can't measure is an assistant you can't trust in production. Kelu says when your sources don't cover a question, collects user feedback, records full traces, and turns weak spots into a concrete content backlog.
The quality loop
Find, inspect, fix, repeat
Honest Answers
- Checks whether your sources cover a question before answering
- Says "I don't know" instead of guessing, with no sources shown
- Numbered citations link to the page and section behind the answer
- Citation coverage shown in analytics
- In a helpdesk, answers below your confidence threshold are never posted
Human Signals
- Per-message feedback across widget, bots, and portal
- Feedback aggregated into quality reporting
- Simulations replay historical tickets before launch
- Human review queues for simulation results
- Deflection counted only for confident answers
Deep Inspection
- Traces for chat, search, MCP, and deflection
- Citations shown in every conversation
- Token counts per answer in traces
- Knowledge-gap report of weak and missing topics
- Doc proposals draft the missing answers
From weak answer to fixed doc
Quality problems in RAG are almost always content problems. Kelu connects the measurement to the fix: unanswered questions and feedback find the weak answers, gap analytics rank what's missing, and doc proposals draft the missing answer — turning quality management into an editorial workflow.
- Feedback and citation coverage track answer quality
- Gap report ranks missing knowledge by question volume
- Doc proposals generate drafts your team reviews and adds to Q&A
- Ask the question again to confirm the fix landed
Conversations
“How do I export audit logs to our SIEM?”
The Acme docs don’t cover exporting audit logs, so I can’t answer that.
Coverage Gaps
Exporting audit logs
9 conversations · 7 people · asked 14 times
Suggestion: Document how admins export the audit log, and in which formats.
Knowledge Maintenance
Q&A
Answers you can check, from a system that won't guess
It says "I don't know"
When your content does not cover a question, Kelu does not guess. In a helpdesk it posts nothing and tags the ticket for your team.
Every answer from your docs shows its sources
Numbered links point to the page and section each part of the answer came from, so anyone can check it in one click.
A bad sync can't empty it
If one sync would remove more than 45% of a source's pages, Kelu holds the removals until a person approves them.
Got questions?
Frequently Asked Questions
Common questions about Answer Quality
Kelu says it doesn't know and shows no sources, instead of guessing. The question is grouped into a topic under Coverage Gaps, so your team sees what to write next.
Yes. Simulations replay historical helpdesk tickets through the assistant and produce reviewable results with a human sign-off step — so you know deflection behavior before enabling it.
Each trace records the question, the generated answer, latency, and token counts — across chat, search, MCP, deflection, and API queries. A search trace also lists the results it returned.
Gap analytics rank unanswered topics by volume, and doc-proposal scans draft candidate answers for those gaps. Your team accepts or dismisses each draft — the AI proposes, humans publish.
Partly. Traces record token counts per answer, and workspace usage metering tracks questions and resolved-versus-deflected outcomes.
Run AI answers like a product
Measure quality from day one — feedback, gaps, and traces are built in.
Free to start · No credit card required