docs(turing): cross-reference Chat, Assets, Token Usage across all entry-point docs

Alexandre Oliveira · claude · Alexandre Oliveira · commit ef919693f77d · 2026-03-20T09:22:09.000-03:00
index.md:
- Restructure into 5 sections: Getting Started, Administration &amp; Development,
  Management (Assets), Generative AI (GenAI, Chat, Token Usage), Technical Reference
- Add all three new pages with descriptions

intro.md:
- Add Assets, Chat, Token Usage to Key Concepts table
- Add 3 new "Where to go next" paths: chat, knowledge base upload, token monitoring

core-concepts.md:
- Knowledge Base section: link to Assets page, describe auto-index on upload/delete
- AI Agents section: link "chat interface" to Chat page
- "Ready to go deeper?" table: add Chat, Assets, Token Usage rows

architecture-overview.md:
- Admin Console module row: list Assets, Chat, Token Usage as named features
- MinIO asset indexing paragraph: rewrite to reference Assets page, add Tika/chunking details
- Frontend tech stack row: update to React + shadcn/ui + Vite, list all UI sections

Co-Authored-By: Claude Sonnet 4.6 &lt;noreply@anthropic.com&gt;
diff --git a/docs-turing/architecture-overview.md b/docs-turing/architecture-overview.md
@@ -84,7 +84,7 @@ The Turing ES application is organized into cohesive modules, each with a well-d
 | **Persistence** | `persistence` | JPA entities, repositories, and DTOs for all domain objects |
 | **Security** | `spring/security` | Keycloak OAuth2 and HTTP Basic authentication |
 | **API Layer** | `api` | REST controllers and GraphQL resolvers exposed to clients |
-| **Admin Console** | React (`turing-react`) | Browser-based UI for administration, configuration, and MinIO file/folder management |
+| **Admin Console** | React (`turing-react`) | Browser-based UI: administration, SN Site configuration, [Assets](./assets.md) file manager (MinIO), [Chat](./chat.md) interface, and [Token Usage](./token-usage.md) dashboard |
 
 ---
 
@@ -142,7 +142,7 @@ sequenceDiagram
 
 **Embedding stores:** If Generative AI is enabled for a Semantic Navigation site, a vector embedding is generated for each indexed document and written to the configured embedding store. Turing ES supports three embedding backends via Spring AI: **ChromaDB**, **PgVector** (PostgreSQL extension), and **Milvus**. Only one is active per deployment. The default embedding store and embedding model are defined globally in **Administration → Settings**.
 
-**MinIO asset indexing:** Turing ES includes a file and folder management UI in the admin console, backed by MinIO as the object storage layer. Files uploaded through this interface are processed by the indexing pipeline and stored as vector embeddings, making them available for RAG queries alongside Semantic Navigation content.
+**MinIO asset indexing:** Turing ES includes an **[Assets](./assets.md)** file manager in the admin console, backed by MinIO as the object storage layer. Files are uploaded via drag-and-drop, organized into folders, and automatically indexed as vector embeddings on upload (and unindexed on deletion). A batch "Train AI with Assets" operation processes all files using Apache Tika for text extraction, chunking at 1,024 characters, and storing embeddings in the active vector store.
 
 **Application logs in MongoDB:** When MongoDB is configured, Turing ES ships with a custom Logback appender that extends `ch.qos.logback`. Every log entry generated by the application — including indexing events, search requests, errors, and system events — is persisted to MongoDB in addition to standard output. These logs are exposed in the admin console, giving administrators full visibility into application behavior without requiring access to the server file system or a separate log management tool.
 
@@ -221,7 +221,7 @@ sequenceDiagram
 | **Load Balancer** | Apache HTTP Server | Optional; required for high-availability cluster deployments |
 | **Connector System** | Viglet Dumont DEP | Separate application; feeds Turing ES via Artemis |
 | **Build System** | Apache Maven | Multi-module project |
-| **Frontend** | React + TypeScript | Admin console (`turing-react`) |
+| **Frontend** | React + TypeScript + shadcn/ui + Vite | Admin console (`turing-react`) — includes SN configuration, [Assets](./assets.md) manager, [Chat](./chat.md) interface, [Token Usage](./token-usage.md) dashboard |
 | **Containerization** | Docker / Docker Compose | Available in `containers/` directory |
 | **Orchestration** | Kubernetes | Manifests available in `k8s/` directory |
 | **API Protocols** | REST + GraphQL | Swagger UI available in development mode |
diff --git a/docs-turing/getting-started/core-concepts.md b/docs-turing/getting-started/core-concepts.md
@@ -128,6 +128,8 @@ This is called **Retrieval-Augmented Generation (RAG)**. The LLM does not make t
 
 Beyond SN Sites, Turing ES provides a **Knowledge Base** — a file and folder interface in the admin console (backed by MinIO) where you can upload documents directly. These files are also indexed as vector embeddings and become available to AI Agents as a searchable knowledge source, independent of any connector or SN Site.
 
+The Knowledge Base is managed through the **[Assets](../assets.md)** page: drag-and-drop upload, folder navigation, inline preview, and batch AI training with real-time progress. Uploaded files are indexed automatically on upload and unindexed on deletion.
+
 ### AI Agents
 
 An **AI Agent** is a named assistant that you compose from three ingredients:
@@ -136,9 +138,9 @@ An **AI Agent** is a named assistant that you compose from three ingredients:
 - A set of **tools** it can call (search your SN Sites, query the Knowledge Base, browse the web, run Python code, get financial data, and more)
 - Optionally, one or more **MCP Servers** — external services that provide additional tools via the Model Context Protocol
 
-Each AI Agent appears as its own tab in the Turing ES chat interface. You can have a "Research Assistant" that combines SN Site search with web browsing, a "Data Analyst" that can run Python code and query your knowledge base, and a "Support Agent" that only sees your product documentation — all on the same platform.
+Each AI Agent appears as its own tab in the **[Chat](../chat.md)** interface. You can have a "Research Assistant" that combines SN Site search with web browsing, a "Data Analyst" that can run Python code and query your knowledge base, and a "Support Agent" that only sees your product documentation — all on the same platform.
 
-See [GenAI & LLM Configuration](../genai-llm.md) for how to set all of this up.
+See [GenAI & LLM Configuration](../genai-llm.md) for how to configure agents, tools, and MCP Servers.
 
 ---
 
@@ -201,6 +203,9 @@ npm install @viglet/turing-sdk
 | Understand the full system architecture | [Architecture Overview](../architecture-overview.md) |
 | Configure Spotlights, Targeting Rules, or Merge Providers | [Semantic Navigation Concepts](../sn-concepts.md) |
 | Set up GenAI, RAG, or AI Agents | [GenAI & LLM Configuration](../genai-llm.md) |
+| Use the AI chat interface | [Chat](../chat.md) |
+| Upload documents to the Knowledge Base | [Assets](../assets.md) |
+| Monitor LLM token consumption | [Token Usage](../token-usage.md) |
 | Secure Turing ES with Keycloak | [Security & Keycloak](../security-keycloak.md) |
 | Install Turing ES | [Installation Guide](../installation-guide.md) |
 
diff --git a/docs-turing/getting-started/intro.md b/docs-turing/getting-started/intro.md
@@ -59,6 +59,9 @@ These are the main building blocks you will work with in Turing ES. You do not n
 | **Merge Providers** | Rules that combine documents from two different connectors into one enriched result. | [Semantic Navigation](../sn-concepts.md) |
 | **RAG** | Retrieval-Augmented Generation — finding relevant documents and using them to ground an LLM's response. | [GenAI & LLM](../genai-llm.md) |
 | **AI Agent** | A named assistant that combines an LLM with a set of tools and knowledge sources. | [GenAI & LLM](../genai-llm.md) |
+| **Assets** | A file manager backed by MinIO where you upload documents to feed the Knowledge Base — files are automatically indexed as vector embeddings. | [Assets](../assets.md) |
+| **Chat** | The conversational AI interface — three tabs for direct LLM chat, Semantic Navigation search, and each configured AI Agent. | [Chat](../chat.md) |
+| **Token Usage** | A dashboard showing LLM token consumption by model, day, and month — useful for monitoring AI costs. | [Token Usage](../token-usage.md) |
 
 ---
 
@@ -78,6 +81,15 @@ Not sure where to start? Here is a suggested path depending on what you want to
 **I want to add generative AI to my search**
 → Read [GenAI & LLM Configuration](../genai-llm.md).
 
+**I want to use the AI chat interface**
+→ Go to [Chat](../chat.md) for direct LLM, Semantic Navigation, and AI Agent tabs.
+
+**I want to upload documents to the Knowledge Base**
+→ Go to [Assets](../assets.md) to manage files and train the RAG index.
+
+**I want to monitor LLM usage and costs**
+→ Go to [Token Usage](../token-usage.md).
+
 **I want to secure Turing ES with SSO**
 → Go directly to [Security & Keycloak](../security-keycloak.md).
 
diff --git a/docs-turing/index.md b/docs-turing/index.md
@@ -12,28 +12,58 @@ slug: /
 - **Current Release:** 2026.1
 - **GitHub:** [https://github.com/openviglet/turing](https://github.com/openviglet/turing)
 
+---
+
 ## Getting Started
 
 | Guide | Description |
-|-------|-------------|
+|---|---|
 | [What is Turing ES?](/turing/getting-started/intro) | Platform overview and key capabilities |
 | [Core Concepts](/turing/getting-started/core-concepts) | SN Sites, connectors, search, GenAI — the mental model |
 | [Installation Guide](/turing/installation-guide) | How to install and configure Turing ES |
 
-## Reference
+---
+
+## Administration & Development
 
 | Guide | Description |
-|-------|-------------|
-| [Administration Guide](/turing/administration-guide) | Console reference: Search Engine, SN Sites, GenAI, Integration |
-| [Developer Guide](/turing/developer-guide) | REST API, SDKs, and integration examples |
-| [Release Notes](/turing/release-notes) | Release history and changelog |
-| [Javadoc](https://turing.viglet.com/latest/javadoc/) | API Reference (external) |
+|---|---|
+| [Administration Guide](/turing/administration-guide) | Console reference: users, roles, API tokens, search engine, SN Sites, global settings |
+| [Developer Guide](/turing/developer-guide) | REST API, SDKs, tech stack, and contribution guide |
+
+---
+
+## Management
+
+| Guide | Description |
+|---|---|
+| [Assets](/turing/assets) | File manager backed by MinIO — upload, organise, and train the RAG Knowledge Base |
+
+---
+
+## Generative AI
+
+| Guide | Description |
+|---|---|
+| [Generative AI & LLM Configuration](/turing/genai-llm) | LLM providers, RAG, embedding models, Tool Callings, MCP Servers, AI Agents |
+| [Chat](/turing/chat) | Chat interface — direct LLM, Semantic Navigation, and AI Agent tabs |
+| [Token Usage](/turing/token-usage) | Monitor and analyse LLM token consumption by model and day |
+
+---
 
 ## Technical Reference
 
 | Guide | Description |
-|-------|-------------|
+|---|---|
 | [Architecture Overview](/turing/architecture-overview) | Component diagram, indexing and search flows, deployment topologies |
 | [Semantic Navigation Concepts](/turing/sn-concepts) | Targeting Rules, Spotlights, Merge Providers, Facets, search response |
-| [Generative AI & LLM Configuration](/turing/genai-llm) | LLM providers, RAG, Tool Callings, MCP Servers, AI Agents |
 | [Security & Keycloak](/turing/security-keycloak) | OAuth2/OIDC setup, Apache reverse proxy, security checklist |
+
+---
+
+## Other
+
+| Guide | Description |
+|---|---|
+| [Release Notes](/turing/release-notes) | Release history and changelog |
+| [Javadoc](https://turing.viglet.com/latest/javadoc/) | Java API Reference (external) |
