Overview

ELS Compass is a private AI compass for enterprise knowledge search. It is part of the Energy Logserver product family.

Business teams use it to ask questions about contracts, procedures, policies, reports and internal documentation, and to get answers that cite the document passages behind them. Documents, indexes and answers stay inside infrastructure the customer controls. Normal product operation does not require an external AI API.

What Problem It Solves

Employees who need an answer from internal documents have two bad options: read through the documents by hand, or paste sensitive content into a public chatbot. The first is slow. The second leaks knowledge the organisation is obliged to protect.

ELS Compass replaces both with a governed knowledge workspace. Documents are ingested into a local index, questions are answered from retrieved evidence, and every answer points back to the passages it used.

Core Capabilities

Capability	What it gives you
Document ingestion	PDF, DOCX, CSV, Markdown and TXT files are parsed, chunked, summarised, embedded and indexed.
Workspaces	Teams, projects, customers and data domains stay separated, with their own documents and conversations.
Hybrid retrieval	Semantic relevance and classic keyword search run against a private ELS/OpenSearch index.
Grounded answers	Responses are generated from retrieved evidence and carry citations to the source passages.
Conversation history	Questions, answers and citations stay scoped to the workspace they belong to.
RBAC	Product access and workspace membership are enforced at the backend API level.

How It Works

The workflow stays simple for business users while the platform handles ingestion, indexing, retrieval, answer generation and permissions underneath.

Connect knowledge. Documents enter a selected workspace and become controlled, searchable business knowledge.
Index privately. Content is parsed, structured, chunked and indexed in a customer-controlled environment.
Ask securely. Users ask in natural language, and the system searches only the knowledge their role permits.
Verify the answer. The response arrives with citations, so a reader can check the evidence immediately.

For what happens inside each step, see Architecture.

Permission Model

Product access and workspace access are two separate questions.

Product entitlements decide whether a person may enter ELS Compass at all:

Entitlement	Meaning
`aibs-access`	May enter the product.
`aibs-admin`	May administer product users, and only when combined with `aibs-access`.

Workspace roles decide what a person may do inside a given workspace:

Role	Scope
Owner	Full control. Deletes workspaces, manages owners, configures access and members.
Admin	Manages workspace state, documents, and non-owner members. Renames and archives workspaces.
Member	Uploads and processes documents, asks questions, uses history and citations.
Viewer	Reads workspace state and asks questions. No file or role management.

A product administrator does not automatically get access to every workspace. Workspace visibility comes from workspace membership. Details are in Security.

Deployment Shape

The API, retrieval index, ingestion workers, parsing, retrieval and answer generation all run inside customer-controlled infrastructure. GPU workers are optional and used for heavier workloads.

Licensing is not metered per token, so a question does not turn into a variable expense.

Where To Go Next

You want to	Read
Install the platform	Installation
Understand the internals	Architecture
Configure a deployment	Configuration
Use the product day to day	User Guide
Run and maintain it	Operations
Integrate with it	API