Real-time text translation

Azure AI Translator converts text between more than 100 languages on request. A Beam agent sends incoming text to the service, reads the returned translation and its confidence, and applies your rules for where the output goes next. Translations that meet your confidence bar are written back to the record or message automatically. Content that returns low confidence, contains sensitive terms, or needs a certified rendering is routed to a person, who reviews the wording before it is used. The agent handles routine language conversion while people check the cases that carry risk.

Batch document translation

Azure AI Translator can translate whole documents in batches while keeping their layout. A Beam agent submits a set of files when your workflow triggers, reads the translated output, and files each result where your process expects it. Batches that complete cleanly within your rules are recorded and passed along without review. Documents that fail to translate, come back with formatting problems, or carry legal or contractual weight are handed to a person before distribution. The agent moves routine batch work through, leaving high-stakes documents for a human to confirm.

Language detection

Azure AI Translator can identify the language of incoming text before translating it. A Beam agent reads new content, uses detection to determine the source language, and applies your rule for how each language should be handled or routed. Text detected in a supported language proceeds to translation and recording automatically. Content where detection is uncertain, returns a mix of languages, or falls outside your supported set is flagged to a person, who decides how to handle it. The agent uses detection to route work, while people resolve the ambiguous cases.