Logo Event-Driven Architecture & Industrial Edge EN ES

AI Playground + Kafka MCP

OpenShift AI 3.3 incluye Gen AI Studio (Playground) para interactuar con LLM desplegados. Combinado con el Kafka MCP Server registrado a través del MCP Gateway de Kuadrant, el Playground se convierte en un agente de IA capaz de consultar datos Kafka en vivo.

Arquitectura

Componente Rol

RHOAI Playground

Interfaz de chat conectada al InferenceService Granite 3.1 2B (KServe)

Granite 3.1 2B Instruct

LLM de huella reducida ejecutándose con llama.cpp en KServe; realiza razonamiento en lenguaje natural

MCP Gateway (Kuadrant)

Proxy Streamable-HTTP que expone los servidores MCP registrados bajo un endpoint unificado

Kafka MCP Server

Servidor FastMCP en Python con 4 herramientas: kafka_list_topics, kafka_describe_topic, kafka_read_messages, kafka_consumer_groups

Kafka Cluster

Clúster CDC (cdc-cluster) en el namespace kafka-cdc con más de 12 topics

Endpoint del MCP Gateway

URL:  https://mcp-gateway.apps.<CLUSTER_DOMAIN>/mcp

Herramientas Kafka disponibles

Herramienta Descripción Parámetros

kafka_list_topics

Lista todos los topics Kafka no internos con el número de particiones

ninguno

kafka_describe_topic

Muestra offsets de partición, recuentos de mensajes y marcas de agua de un topic

topic_name (cadena)

kafka_read_messages

Lee los N mensajes más recientes de un topic (máx. 20)

topic_name (cadena), count (entero, predeterminado 5)

kafka_consumer_groups

Lista todos los grupos de consumidores con sus estados

ninguno

Topics Kafka disponibles

Topic Contenido

cv.face.detections

Eventos NeuroFace CV: person_registered, person_detected, ovms_model_status

cdc.public.customers

Eventos de cambio CDC de la tabla customers

cdc.public.orders

Eventos de cambio CDC de la tabla orders

bpm.approval.requests

Eventos del flujo de aprobación BPM desde Kogito/SonataFlow

bpm.onboarding.metrics

Métricas de onboarding BPM y telemetría de procesos

dlq.cdc-errors

Cola de mensajes fallidos para eventos CDC erróneos

dlq.cv-errors

Cola de mensajes fallidos para eventos del pipeline CV erróneos

dlq.cdc-camel-errors

Cola de mensajes fallidos para integraciones Camel K erróneas

Prompts del Playground

Prompt 1 — Visión general del inventario

List all Kafka topics available in the cluster.
For each topic, show the name and number of partitions.
Highlight any topics related to face detection or computer vision.

Prompt 2 — Análisis del pipeline CV

Read the last 10 messages from the topic cv.face.detections.
Summarize:
- How many unique people were detected?
- What detection methods are being used?
- What is the average confidence score?
- Are there any OVMS model status events? If so, is the model healthy?

Prompt 3 — Inspección de eventos CDC

Describe the topic cdc.public.orders and tell me how many total messages it has.
Then read the last 5 messages and summarize the order data:
- What fields are in each record?
- Are these INSERT, UPDATE, or DELETE operations?
- Show a brief table with customer_id, amount, and operation type.

Prompt 4 — Estado de los grupos de consumidores

List all consumer groups in the Kafka cluster.
For each group, show the state (Stable, Empty, Dead, etc.).
Identify any consumer groups that might indicate a problem
(e.g., Dead or with rebalancing state).

Prompt 5 — Análisis de colas de mensajes fallidos

Check the following dead-letter queue topics:
- dlq.cdc-errors
- dlq.cv-errors
- dlq.cdc-camel-errors

For each one, tell me how many messages are in it.
If there are messages, read the latest 3 and summarize
what errors occurred and which components generated them.

Prompt 6 — Informe de detección facial de seguridad

Read the last 20 messages from cv.face.detections.
Generate a security report with:
1. Timeline of detections (person name + timestamp)
2. Total unique identities currently registered
3. Detection frequency (events per minute estimate)
4. Model server health status
Format the report as a professional security briefing.

Prompt 7 — Eventos de flujos BPM

Describe the topics bpm.approval.requests and bpm.onboarding.metrics.
Read the last 5 messages from each and explain:
- What business processes are generating events?
- What is the approval workflow doing?
- Are there any onboarding metrics available?

Prompt 8 — Estado de topics a nivel de clúster

List all topics and describe each one that has more than 0 messages.
Create a dashboard-style summary:
- Topic name
- Total messages
- Number of partitions
- Category (CDC / CV / BPM / DLQ / Internal)
Flag any topic that looks unhealthy (e.g., too many DLQ messages
relative to the source topic).

Prompt 9 — Pipeline NeuroFace de extremo a extremo

I want to understand the NeuroFace CV pipeline end-to-end.
1. Read the latest messages from cv.face.detections
2. Check if there are any errors in dlq.cv-errors
3. Describe the cv.face.detections topic to see throughput

Based on this data, explain:
- Is the pipeline healthy?
- How many events are being generated per hour (estimate)?
- Are any detections failing?

Prompt 10 — Correlación entre sistemas

Read the last 5 messages from these topics:
- cv.face.detections
- cdc.public.orders
- bpm.approval.requests

Find any correlations:
- Are detections happening around the same time as orders?
- Is there any pattern between face detection events and BPM workflows?
Provide a timeline visualization in text format.

Referencias