Messagebird NamedEntityRecognition

Name: NamedEntityRecognition API
Brand: NamedEntityRecognition
Availability: InStock

★ Only Publicly Available OpenAPI DocumentAI/MLMl InferenceapiKey2 Endpoints

Install Jentic One Beta

Connect the NamedEntityRecognition to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the NamedEntityRecognition, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with NamedEntityRecognition API.

Extract dates and locations from a customer SMS to pre-fill an appointment booking form

Pull person names and organisations out of inbound chat replies to enrich a CRM record

Identify monetary amounts mentioned in a billing question before routing it to finance

GET STARTED

Start building with NamedEntityRecognition API

Explore with Jentic

View OpenAPI Document

For Agents

Extract structured entities (names, places, dates, organisations) from a customer message so an agent can fill slots without keyword parsing.

Use for: Extract named entities from this customer message, Pull dates and locations out of an inbound SMS, Identify person and organisation names in a chat reply, Check whether NER supports German

Not supported: Does not handle intent classification, language detection, or message delivery - use for named entity extraction from short text only.

Jentic publishes the only available OpenAPI specification for NamedEntityRecognition, keeping it validated and agent-ready. The MessageBird Named Entity Recognition API extracts structured entities (names, locations, dates, organisations) from short text inputs, turning unstructured customer messages into machine-readable slots. The predict-entities endpoint accepts a phrase and returns each detected entity with its type, while the supported-languages endpoint indicates which locales the model recognises. NER is the natural pairing for intent classification when an agent needs both 'what does the customer want' and 'what specific values did they mention'.

Use Cases

Patterns agents use NamedEntityRecognition API for, with concrete tasks.

★ Appointment Booking Slot Filling

When a customer texts 'Can I move my appointment to Friday at the Dublin clinic?', NER extracts the date and the location as structured entities. The agent can then call the calendar API with real values instead of trying to parse the phrase by hand. This makes natural-language booking flows reliable enough to deploy without a human in the loop.

POST the booking message to /api/v1/predict-entities and use the returned date and location entities to call the calendar booking endpoint

CRM Record Enrichment

Inbound free-text messages from leads often mention company names, contact people, and locations. NER turns these mentions into structured fields the CRM can consume, so an agent can update a lead record without the user filling out a form. Entity types are returned alongside values so the integration knows which CRM property to update.

Call POST /api/v1/predict-entities on a sales reply and update the CRM contact's company and city fields with the returned entity values

Compliance Redaction

When recording or analysing customer messages, agents need to redact personally identifying information. NER identifies names, locations, and other PII candidates so the redaction step can mask them before storage or model training. Pair this with a deny list to suppress sensitive entity types entirely.

Run POST /api/v1/predict-entities on the message, replace each PERSON or LOCATION entity with a placeholder, and store the redacted text

AI Agent Integration via Jentic

An AI agent uses Jentic to discover NER by intent search ('extract entities from a message'). Jentic returns the predict-entities operation with its input schema, so the agent submits a phrase and receives structured entity records in one load-and-execute step. The MessageBird AccessKey lives in your Jentic One instance and never enters the agent's context.

Use Jentic to search 'extract entities from a phrase', load the predict-entities schema, and execute it to populate the slots of a downstream booking action

Key Endpoints

2 endpoints — jentic publishes the only available openapi specification for namedentityrecognition, keeping it validated and agent-ready.

METHOD

PATH

DESCRIPTION

POST

/api/v1/predict-entities

Extract named entities from a submitted phrase

GET

/api/v1/supported-languages

List supported NER languages

POST

/api/v1/predict-entities

Extract named entities from a submitted phrase

GET

/api/v1/supported-languages

List supported NER languages

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

Your MessageBird AccessKey is stored once, encrypted, by your own Jentic One instance and attached at execution time. It never enters the agent's prompt, logs, or context.

Intent-based discovery

Agents search Jentic by intent such as 'extract entities from a phrase', and Jentic returns the predict-entities operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

Time to first call

FAQs

Specific to using NamedEntityRecognition API through Jentic.

Why is there no official OpenAPI spec for NamedEntityRecognition?

MessageBird does not publish an OpenAPI specification for the NER API. Jentic generates and maintains this spec so that AI agents and developers can call it via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

What authentication does the MessageBird NER API use?

An API key in the `Authorization` header, prefixed with the literal word `AccessKey`. Through Jentic the key sits in your Jentic One instance and is injected at request time, so the agent never sees the raw value.

Which entity types does the model return?

The predict-entities response surfaces standard entity categories (such as person, location, organisation, and date) for the submitted phrase. Inspect a few sample responses to confirm the exact label set for your traffic before relying on a specific type.

What languages does NER support?

Call GET /api/v1/supported-languages to fetch the live list of supported locales rather than hardcoding it. The set evolves as MessageBird extends coverage.

How do I extract entities through Jentic?

Search Jentic for `extract entities from a phrase`, load the POST /api/v1/predict-entities operation, and execute it with your text. Jentic returns the entities as structured records the agent can iterate over.

Are there published rate limits?

The OpenAPI spec does not declare quantitative rate limits. Check your MessageBird account dashboard or contact MessageBird support for the limits attached to your plan before high-volume usage.

Can I limit what my agent is allowed to do with the MessageBird NER API?

Yes. Because you run Jentic One yourself, your own rules decide which operations and credentials the agent may use. The MessageBird NER API only exposes two operations, so you can grant your agent POST /api/v1/predict-entities to extract entities from a phrase and GET /api/v1/supported-languages to check locale coverage, while leaving out anything else. Since the API takes text in the request body rather than as a URL path resource, scoping the agent to just these two operations is enough to bound exactly what it can call.