Streaming Frontiers - S01E03 - The MCP Expedition Continues

Join Viktor Gamov, Principal Developer Advocate, for Part 2 of his exploration of all things MCP. Viktor believes this protocol will significantly alter the workflow of data streaming engineers. His exploration is based around Confluent’s MCP server, which lets you call Confluent resources using natural language within an MCP client (e.g., Claude Desktop, Windsurf). With it, you can perform all of the Confluent basics, such as: - Managing Kafka topics - Executing Flink SQL - Adding connectors In the episode, Viktor explores adding non-Confluent MCP servers to your results mix. Tune in as he guides you through it. Timecodes: 00:00 - timer starts 03:20 - intro and welcome 06:13 - Current 2025 London updates 13:25 - use code L-PRM-Frontiers when registering for Current London 14:00 - a quick recap from the previous episode 20:20 - using MCP Inspector with mcp-confluent 27:30 - trying Github MCP Server 32:40 - trying Confluent MCP Server 33:40 - creating Python producer for Confluent Cloud 46:00 - creating Flink statements using Confluent MCP 1:01:49 - create Avro schema and attach it to the topic using Confluent Extension 1:05:00 - continue debugging Flink statements 1:10:35 - create Flink views in Confluent Cloud 1:12:58 - it's a wrap!