Use Cases
This section answers one question: what can you do with DIAL, and why is it worth doing on DIAL rather than another way? Each page describes a real-world scenario, explains the value, then shows how you would build it and where to start. It is for evaluators deciding whether DIAL fits a problem, and for builders looking for a concrete starting point.
Every scenario leads with the value and the reasoning, then drops into a "How you'd build it" section with the DIAL components involved and links into the rest of the documentation. You can read the first half to decide, and the second half to act.
Scenarios
- Org-wide AI assistant — give every employee safe, governed access to many models through one interface.
- Chat with your knowledge base — ground answers in your own documents with retrieval-augmented generation.
- No-code AI tools for business teams — let non-developers compose models, tools, and agents without writing code.
- One API for every model provider — call any model through a single interface and swap providers without rewriting apps.
- Centralized cost and usage governance — see, limit, and attribute AI spend across teams and applications.
- Custom production AI agents — build, deploy, and operate bespoke agents with full control over logic and dependencies.
- Embed AI in existing tools — bring DIAL into Microsoft Teams, Excel, automation pipelines, and your own web apps.
- Measure and improve AI quality — make model and prompt changes against metrics instead of intuition.
- Compliance and PII-safe AI — enforce redaction, audit, and access boundaries on every request.
Find your scenario by capability
If you know which DIAL capability you need, this table points you to the scenario where it does the most work.
| If you need… | Start with |
|---|---|
| A governed chat interface for staff | Org-wide AI assistant |
| Answers grounded in your documents | Chat with your knowledge base |
| No-code app and tool composition | No-code AI tools for business teams |
| Provider independence and routing | One API for every model provider |
| Cost limits, quotas, and attribution | Centralized cost and usage governance |
| Custom code, dependencies, and scaling | Custom production AI agents |
| AI inside the tools people already use | Embed AI in existing tools |
| Evaluation and quality measurement | Measure and improve AI quality |
| Redaction, audit, and access control | Compliance and PII-safe AI |