Welcome

Your knowledge base at a glance.

-
Indexed chunks
-
Project
-
Your role

Recent activity

Time (UTC)QuestionTokens
Loading...

Change your password

Ask

Grounded, cited answers from your knowledge base.

Indexing

Upload documents (PDF, Word, and more) or paste text - extracted, chunked, embedded, and stored.

Drop file(s) here, or click to choose - you can select multiple
Upload
/v1/ingest
Extract
loader
Chunk
+ overlap
Embed
vector
Vector DB
persist

Paste text

Documents

Everything indexed in your knowledge base - by type, category, and file.

By file type

TypeDocsChunks

By category

CategorySubcategoryDocsChunks

All documents

FileTypeCategoryNamespaceChunksIndex entriesIndexed (UTC)

Pipeline

Run the full orchestration and watch every stage in real time.

Orchestration stages

Persona lookup (live state)idle
Plan - select sourcesidle
Retrieve - parallel fan-outidle
Fuse - RRF rankingidle
Governance - redact / requireidle
Build promptidle
LLM - generate answeridle

AI Agent

Chat with your knowledge base - and the details to integrate it into your own app.

Integration details

Shown once - copy and store it. Only the hash is kept.

Call it from your app


        

Chat

Scorecard

The health of your knowledge base - answer quality, coverage gaps, speed, adoption, and spend.

Answer quality
-
Grounded answers
have ≥1 citation
-
Satisfaction
thumbs up / total
-
Coverage gaps
questions with no match
-
Questions asked
recent window
Speed & reliability
-
Avg latency
end-to-end
-
p95 latency
slowest 5%
-
Open failures
need attention
-
Active users
of provisioned
Cost & knowledge base
-
Spend
estimated, recent
-
Cost / question
avg
-
Documents
chunks indexed
-
Stale documents
not re-indexed in 90d

Unanswered questions

Questions that retrieved nothing - the content your knowledge base is missing.

  • Loading...

Knowledge by category

Loading...

Cost Intelligence

Who spends, on what - plus ML forecasts and a per-question cost estimate from your own history.

Spend forecast
-
Projected spend
next 30 days
-
Likely range
95% band
-
Avg / day
recent
-
Trend
direction

Estimate a question's cost before you ask it

Predicted from the most similar questions in your own usage history (embeddings + nearest-neighbour).

Top spenders

UserRequestsTokensCost

Cost by model

ModelRequestsTokensCost

Cost by category - which parts of your knowledge base drive spend

CategoryRequestsTokensCost

Daily spend (last 30 days)

Loading...

Cost spikes

Days whose spend was anomalously high (robust z-score).

  • -

Expensive question types

Clusters of similar questions, by average cost.

-

Usage & Billing

Every request's tokens and estimated cost - for your project only.

Recent requests

When you use your own LLM key/resource, these numbers reconcile 1:1 with your provider's billing dashboard.

Time (UTC)QuestionModelInOutTotalCost (USD)

Settings

Configure your LLM and choose how your API key is handled.

Key handling - how your LLM key is stored

Active engine: -

Document storage - where your document vectors live

Keep your documents in our managed store (isolated to your project), or bring your own database so they never touch our system.

Current: -

About CheriMind

Your organization's knowledge, answered - grounded, cited, and under your control.

What is CheriMind?

CheriMind turns your own documents into a trustworthy AI assistant. Upload your files - policies, FAQs, contracts, manuals - and anyone on your team can ask questions in plain language and get clear answers built only from your content, with the exact sources shown for every answer. No coding, no data science, no guessing.

Our vision

Every organization should be able to ask its own knowledge a question in plain language - and get a trustworthy, sourced answer - without handing its data or its AI keys to a black box.

Our mission

Make grounded, trusted AI over your own documents effortless and safe: full transparency (every answer cites its sources), full control (your data, your AI keys, even your own database), and zero engineering.

How CheriMind helps you

Answers, not searching

Ask in plain English and get a direct answer from your documents - not a pile of links to read.

Trust every answer

Each answer shows the exact source passages it used, so you can verify it. No made-up facts.

No engineering

Upload a file, ask a question. Your team is self-sufficient - no developers required.

Costs under control

See who spends what, forecast next month, set budgets, and get alerts before you overspend.

Built-in governance

Sensitive details like emails and phone numbers are hidden automatically, and every query is logged.

Your teams, separated

HR, Finance, Support and more stay organised and isolated - no cross-mixing of data.

What makes CheriMind different
  • Grounded & cited by design - a glass box, not a black box. Every answer is traceable to your sources.
  • Your data stays yours - keep it in our isolated managed store, or bring your own database so it never leaves your systems.
  • Your AI key, your rules - four key-handling modes, from fully managed to "we never see your key or your answer".
  • Cost Intelligence - spend forecasts, per-question cost prediction from your own history, and spike alerts - rare in this category.
  • Any document, any domain - PDF, Word, Excel, images (OCR) and more, for any industry.
  • Zero-code onboarding - add a new team or client as a project - no rebuild, no developers.
What each page does
PageWhat it is for
OverviewSnapshot of your knowledge base, your role, and recent activity.
AskAsk a question; get a grounded answer with the exact sources it used.
IndexingAdd documents (upload a file) or paste text into your knowledge base.
DocumentsSee and manage everything you've added, by type and category.
PipelineWatch, step by step, exactly how an answer is built.
AI AgentChat with your knowledge base, plus the details to call it from your own app.
ScorecardAnswer quality, coverage gaps, speed, and spend at a glance.
Cost IntelligenceForecast spend, set budgets, and price a question before you ask it.
Usage & BillingLine-by-line record of every question, its tokens, and cost.
SettingsConnect your AI model, choose how your key is handled, and where your data is stored.

Open the round ? button on any page for a detailed, page-specific guide with examples.

Getting started
  1. Indexing - upload your documents (or paste text).
  2. Ask - type a question; get a grounded, cited answer.
  3. Settings - connect your own AI model and choose where your data is stored.
  4. Scorecard & Cost Intelligence - track answer quality, coverage gaps and spend.

Tip: the round ? button (bottom-right) gives help specific to whichever page you're on.