> ## Documentation Index
> Fetch the complete documentation index at: https://handbook.aiql.io/llms.txt
> Use this file to discover all available pages before exploring further.

# About AiQL

> What AiQL is and why it exists

AiQL combines symbolic graph representation, deep learning, and human-in-the-loop verification into a platform for knowledge-intensive operations. It is aimed at healthcare, finance, government, and other regulated industries where answers must be precise, traceable, and defensible.

The platform sits as a reasoning layer over existing organizational data: databases, enterprise systems, documents and policies, expert knowledge, and external standards. What customers use is described under [Products](/products/overview). Ingestion and query sit on the [Middleware API](/products/middleware). The parent company is in Singapore, with subsidiaries in the United States and Saudi Arabia. See [Legal](/company/legal).

## How AiQL differs from generic AI

Generic systems search documents and guess. AiQL reasons across a structured knowledge layer that the organization maintains.

* Answers include step-by-step reasoning chains and audit trails, rather than a black-box output.
* Accuracy depends on encoded organizational knowledge, not on which model is called.
* As structured knowledge grows, answers become more precise. Model-only systems can lose precision as information volume grows.
* Inference cost does not have to rise with every query the way usage-priced model calls do. Value is meant to compound in the knowledge layer instead.

## How knowledge compounds

<Steps>
  <Step title="Experts sculpt knowledge">
    Business users contribute rules, processes, decisions, and judgment.
  </Step>

  <Step title="Knowledge compounds">
    Each use refines the knowledge layer. Institutional intelligence builds.
  </Step>

  <Step title="Reasoning improves">
    More structured knowledge produces more precise, explainable answers.
  </Step>

  <Step title="The organization adopts">
    Better answers increase daily use, which feeds the cycle.
  </Step>
</Steps>

AI model outputs can depreciate as information volume increases. AiQL is designed so that as organizational knowledge is scaled and structured, the system becomes more capable.

## Where it is aimed

These are target segments, not named clients. Client profiles and geographic focus are on [Market](/market/overview). Named traction belongs on [Clients](/clients/overview).

<Columns cols={2}>
  <Card title="Healthcare" icon="heart-pulse">
    Treatment protocols across standards, insurance, and regulations. Clinical reasoning that can be applied beyond a single physician.
  </Card>

  <Card title="Financial services" icon="landmark">
    Credit decisioning with explainable rationale chains. Regulatory interpretation that is auditable and consistent. Fraud pattern reasoning with traceable logic.
  </Card>

  <Card title="Pharma and life sciences" icon="flask">
    Drug interaction reasoning across layered regulatory requirements. Pharmacovigilance with decision traceability. Regulatory submissions supported by structured guidelines and protocols.
  </Card>

  <Card title="Construction and real estate" icon="building">
    Contract interpretation across FIDIC, SCA, and local regulations, with documented reasoning for permit decisions. Design specifications that reuse the organization's expertise and past work.
  </Card>
</Columns>