AI Memory That Actually Works

Phil Laureano, a familiar face at Alt.Net, will be talking about new AI Memory techniques that will change how your AI coding sessions work.

Every AI coding session starts the same: re-explain your codebase, watch the context window fill up, lose everything when it compacts. Repeat tomorrow. Phil got tired of the loop. He forked an open-source coding harness and bolted on a persistent memory system using patterns most .NET devs already know — pub/sub, compression hierarchies, topic-based messaging. Now his sessions run for days without losing context, agents share knowledge across terminals instantly, and the system can turn a ramble into a finished song.

This talk traces the evolution from Obsidian markdown files to a full memory architecture, each friction point driving the next level. Live demos woven throughout — no slides, two terminals, shared memory, and you'll hear what AI memory sounds like when it works.

As per our long standing tradition, we will be giving away 2 JetBrains licenses at the end of the talk

Stream at: https://www.twitch.tv/sydneyaltnet