Nurai transforms your documents, policies, and institutional memory into a verifiable knowledge layer — deployed natively on WhatsApp, LINE, and the web. Every answer is cited. Every gap is logged.
Organisations accumulate deep institutional knowledge over years — in manuals, meeting notes, policies, and email threads. The gap is not the content. It is the structured, auditable path from question to verified answer.
Generic LLMs generate plausible-sounding answers that are not anchored to your actual documents. In regulated or high-stakes contexts, that is a liability, not a feature.
When a question cannot be answered, most systems either fabricate or fail silently. Neither tells you what your corpus is missing.
Nurai does not treat all answers as equal. Every response is classified into one of three transparent states — so users always know what they are getting, and managers always know what is missing.
The answer was found verbatim or near-verbatim in the indexed knowledge base. The response includes the exact source document and section reference, making it fully auditable.
Source linked — no interpretation added. Confidence above retrieval threshold.
The answer draws on multiple chunks from different documents. The model connects them explicitly, not by assumption. All contributing sources are listed.
Multi-source references listed. Reasoning is transparent and traceable.
No relevant source was found above the confidence threshold. The user receives a transparent "not in our current knowledge base" response. The query is logged to the Gap Queue for content owners to act on.
Query logged to Gap Queue. Ops team notified. Corpus can be updated.
From document upload to cited answer — a transparent, auditable pipeline with no black-box steps.
Nurai is not a generic chatbot. It is a governed knowledge infrastructure designed for organisations where accuracy, auditability, and access control are non-negotiable.
Every Retrieved or Synthesised answer includes the exact source document, section, and update date. Compliance teams can audit any conversation end-to-end.
Each Advisor can be scoped to internal staff, external users, or specific departments. Sensitive documents are restricted at the corpus level, not the prompt level.
Unanswerable questions are automatically logged with frequency counts. Content owners see exactly what to document next, closing the knowledge loop over time.
Advisors can be configured per language. Thai, English, and regional variants supported. Automatic language detection routes queries without user configuration.
Answer leave, payroll, and benefits questions 24/7. Reduce HR inbox volume by routing routine queries to a verified, citable source.
Distribute regulatory updates and compliance requirements without relying on one-off email chains. Every answer is version-controlled.
Nurai deploys on the platforms your staff and customers already use — no new app, no onboarding friction. One Advisor can serve multiple channels simultaneously.
All plans include the three-state model, source citations, and gap logging. Higher tiers unlock more Advisors, larger corpora, and enterprise governance controls.
One Advisor, one channel, up to 50 source documents. Ideal for a single department pilot.
Up to 5 Advisors, all channels, 500 documents. The right tier for multi-department rollout.
Unlimited Advisors, unlimited documents, on-premise option, SLA, and dedicated success manager.
All prices exclude VAT. Annual billing available at 15% discount. Contact us for academic or NGO pricing.
We run a structured 30-day pilot — live corpus, real channel, measured gap reduction. Most teams see query coverage above 80% before the pilot ends.
Real corpus, real channel, real staff queries. Measurable corpus coverage and gap reduction by day 30.
Knowledge Gaps are declared and logged, not invented. Your compliance and legal teams can verify every response end-to-end.
Connect your drive, configure your Advisor, go live on WhatsApp or LINE within a week. No integration sprint required.