Unified API connections

Nango gives one consistent interface over many third-party APIs, so connected accounts share a common shape. A Beam agent reads records through that interface, such as contacts or tickets from a linked provider, and writes updates back using the same connection. It applies the rules your team approves before any write reaches the downstream tool. When a provider returns an unexpected field, a permission gap, or a record the agent cannot confidently place, it pauses the write and notifies an engineer with the raw response, so a person decides how to proceed rather than guessing.

Scheduled record refreshes

Nango keeps connected data current by fetching records from each provider on a schedule. A Beam agent reads the refreshed set, compares it against the last known state, and updates the matching internal records when the change fits an approved rule. Small, expected differences apply on their own. A large deletion, a schema shift, or a conflict between two sources goes to an engineer, who keeps the decision while the agent presents both versions and the timestamps. This keeps downstream tools working from current data without a person watching every refresh cycle.

Webhook event handling

Nango forwards provider webhooks so downstream systems learn about changes as they happen. When an event arrives, such as a new order or an updated contact, a Beam agent reads the payload, validates it against your rules, and writes the resulting update or notifies the owning team. Routine events flow through untouched. An event with a missing reference, a signature that fails validation, or a type the agent has no rule for routes to an engineer for a decision. The agent notes what it did with each event so the history stays clear.