Browsing Category

MCP

3 posts
Agentic AI with Apache Kafka as Event Broker Combined with MCP and A2A Protocol
Read More
  • 5.7K views
  • 7 minute read

Agentic AI with the Agent2Agent Protocol (A2A) and MCP using Apache Kafka as Event Broker

Agentic AI is emerging as a powerful pattern for building autonomous, intelligent, and collaborative systems. To move beyond isolated models and task-based automation, enterprises need a scalable integration architecture that supports real-time interaction, coordination, and decision-making across agents and services. This blog explores how the combination of Apache Kafka, Model Context Protocol (MCP), and Google’s Agent2Agent (A2A) protocol forms the foundation for Agentic AI in production. By replacing point-to-point APIs with event-driven communication as the integration layer, enterprises can achieve decoupling, flexibility, and observability—unlocking the full potential of AI agents in modern enterprise environments.
Read More
0
0
1
Data Streaming Lake Warehouse and Lakehouse with Confluent Databricks Snowflake using Iceberg and Tableflow Delta Lake
Read More
  • 1.9K views
  • 12 minute read

Databricks and Confluent Leading Data and AI Architectures – What About Snowflake, BigQuery, and Friends?

Confluent, Databricks, and Snowflake are trusted by thousands of enterprises to power critical workloads—each with a distinct focus: real-time streaming, large-scale analytics, and governed data sharing. Many customers use them in combination to build flexible, intelligent data architectures. This blog highlights how Erste Bank uses Confluent and Databricks to enable generative AI in customer service, while Siemens combines Confluent and Snowflake to optimize manufacturing and healthcare with a shift-left approach. Together, these examples show how a streaming-first foundation drives speed, scalability, and innovation across industries.
Read More
0
0
0
Event-Driven Agentic AI with Data Streaming using Apache Kafka and Flink
Read More
  • 8.4K views
  • 12 minute read

How Apache Kafka and Flink Power Event-Driven Agentic AI in Real Time

Agentic AI marks a major evolution in artificial intelligence—shifting from passive analytics to autonomous, goal-driven systems capable of planning and executing complex tasks in real time. To function effectively, these intelligent agents require immediate access to consistent, trustworthy data. Traditional batch processing architectures fall short of this need, introducing delays, data staleness, and rigid workflows. This blog post explores why event-driven architecture (EDA)—powered by Apache Kafka and Apache Flink—is essential for building scalable, reliable, and adaptive AI systems. It introduces key concepts such as Model Context Protocol (MCP) and Google’s Agent-to-Agent (A2A) protocol, which are redefining interoperability and context management in multi-agent environments. Real-world use cases from finance, healthcare, manufacturing, and more illustrate how Kafka and Flink provide the real-time backbone needed for production-grade Agentic AI. The post also highlights why popular frameworks like LangChain and LlamaIndex must be complemented by robust streaming infrastructure to support stateful, event-driven AI at scale.
Read More
0
0
0