BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//hackathonmaxxing//EN
CALSCALE:GREGORIAN
X-WR-CALNAME:Hackathons in Barcelona
BEGIN:VEVENT
UID:meetup:317286539@hackathonmaxxing
DTSTAMP:20261122T163405Z
DTSTART:20261126T171500Z
DTEND:20261126T201500Z
SUMMARY:Network data meets AI: Gigamon\, MCP and LLM evals in Elastic
URL:https://www.meetup.com/barcelona-elastic-fantastics/events/317286539/?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=barcelona
LOCATION:Betahaus | Coworking Barcelona\, Carrer de Vilafranca\, 7\, Gràcia\, 08024 Barcelona\, Spain\, Barcelona
DESCRIPTION:Las charlas se darán en inglés.\n\nBarcelona calling! Join Elastic and Gigamon for two technical talks where network data meets AI: Gigamon network traffic flowing into Elastic and investigated with AI through MCP\, and how to evaluate LLM applications properly with @kbn/evals. Q&A and networking over food and drinks to follow.\n\nAddress:\nBetahaus | Coworking Barcelona\, Carrer de Vilafranca\, 7\, Gràcia\, 08024 Barcelona\, Spain\n\nAgenda:\n18:15 Doors open\n18:30 Talk #1 From packets to answers: Gigamon network data in Elastic\, investigated with AI by Ferran Orsola \\- Principal Sales Engineer at Gigamon\nAbstract: A walkthrough of the Gigamon setup\, followed by a live demo of Gigamon network data flowing into Elastic. We’ll explore the data and dashboards in Kibana\, then show how Gigamon Insights answers questions about that data through its MCP connection to Elastic.\n19:00 Q&A + Short break\n19:20 Talk #2 Evals: Datasets\, Evaluators and Statistical Comparison (with @kbn/evals) by Arturo Lidueña \\- Software Engineer at Elastic\nAbstract: How to evaluate LLM applications\\\, from dataset design to statistical comparison of results\\. Building blocks: datasets and examples\\\, the task under test\\\, and evaluators \\(deterministic checks vs\\. LLM\\-as\\-a\\-judge\\)\\. Online/offline evals\\. Trace\\-based evaluators\\. Statistical analysis: how to select a statistical test for each metric type \\(binary\\\, bounded\\\, unbounded\\) and decide whether a difference is significant\\. At Elastic\\\, we develop @kbn/evals to evaluate Kibana's own AI systems\\. It runs various evaluation suites \\(Agent Builder agents\\\, ES\\|QL generation\\\, alert triage\\\, attack discovery\\) on every pull request across multiple models\\\, with scores and traces stored in Elasticsearch for comparison\\.\n19:50 Q&A\n20:10 Networking with food and drinks\n21:00 Wrap up\n\nThank you Betahaus Barcelona for hosting us and Gigamon for partnering up with us!\n\nIf you’re a longtime Elastic user or just starting your journey\, this is the perfect opportunity to share ideas\, meet new people\, and get inspired.\nInvite your friends and RSVP on this page!\n\n⚡️ Interested in giving a talk? ⚡️\nHave you ever considered presenting on your Elastic use case? We welcome 5-10 minute lightning talks\, 45-minute deep dives\, and everything in between. If you're interested\, please submit via our CFP or send us an email at meetups+emea@elastic.co https://www.meetup.com/barcelona-elastic-fantastics/events/317286539/?utm_source=hackathonmaxxing&utm_medium=referral&utm_campaign=barcelona
END:VEVENT
END:VCALENDAR
