Skip to main content

Apps and agents showcase

Real applications built on DIAL — from retrieval-augmented generation to multi-agent assistants and no-code apps. Each demo shows a finished capability, not how to build it.

DIAL RAG​

DIAL's retrieval-augmented generation: ask questions over your own PDFs, Office files, and text, with answers grounded in the retrieved content.

DIAL Web RAG​

RAG extended to the open web — combining cloud search with DIAL RAG, an example of DIAL's agentic approach where one app uses another as a building block.

ChatHub​

A multi-agent orchestrator that routes each prompt to the right tool — text, image, vision, code interpreter, or RAG — behind a single chat entry point.

ChatHub 2.0 introduces a prompt routing engine that picks the best tool for each request and chains agents for efficient, combined responses.

Code Apps​

Write, deploy, and run Python applications directly in the DIAL UI — no separate hosting to manage — demonstrated by building a simple RAG app end to end.

Quick Apps​

The no-code agent orchestrator: combine MCP toolsets, AI agents, and models into multi-agent workflows without writing code.

A Quick App that returns weather information and clothing recommendations for any city.

Using attached files as a RAG source inside a Quick App.

MCP integration​

Connect DIAL to external systems through the Model Context Protocol, surfaced through the Admin Panel and Quick Apps.

StatGPT​

An open-source "talk to your data" platform for official statistics, built on DIAL — querying, analyzing, and visualizing statistical data through natural language with an agentic architecture.

OMICS AI Assistant​

A multi-agent application for scientific data: specialized agents handle query, visualization, and analysis over an OMICS dataset, with results shareable across a team.

Talk-to-your-data apps​

Two case studies built on DIAL's talk-to-your-data framework, where dedicated agents handle each step behind a simple natural-language interface.

The Athlete Profile Generator builds a detailed profile from just a player's name.

The Oil and Gas AI Assistant explores production data, generating code for visualizations, calculations, and interpretation.

DataViz App​

A code walkthrough of the agent-orchestration app that visualizes and analyzes financial data — the precursor to StatGPT.

Entity Extractor​

Identify and pull predefined information from unstructured documents — for example, a claims-processing assistant that extracts, validates, and reports on insurance claim data.

TimeBase AI Assistant​

A multi-agent assistant that retrieves, analyzes, and monitors time-series financial data in real time — generating indicators, trading signals, and live alerts, with each agent's reasoning on screen.

Guided Conversations​

An interactive-survey application built on DIAL that walks a user through a guided, structured conversation.

Further reading​