About this Event
Hot-Swapping Reality: JIT Context Delivery and the Evolution of the Brain
with Amnesia Labs founder sk!tz0
Most AI systems struggle with 'long-term' memory, often retrieving too much noise or not enough signal. This talk introduces the brain from Amnesia Labs, a system built to solve the context gap through Just-In-Time delivery.
They either pull back too much context and drown in noise, or too little and miss what matters. The result is brittle systems that lose track of what’s important the moment things get complex.
This talk introduces The Brain from Amnesia Labs, a system designed to close that gap.
Instead of static memory or brute-force retrieval, it uses Just-In-Time context delivery, bringing in the right information at the moment it’s needed, and nothing more. Not everything all at once. Not a guess. Precisely what’s required to move forward.
This isn’t just about improving LLM performance. It’s about changing how we think about memory, context, and intelligence in AI systems.
If you’ve worked with embeddings, RAG pipelines, or tried to scale anything beyond a toy system, you’ve seen the limitations firsthand. This is a different approach.
The goal is simple: Less noise. More signal. Systems that actually keep track of what matters.
Hackers Guild AI/LLM Talks is a series of focused, technical sessions where builders share what they’re actually working on in AI. The emphasis is on real systems, real constraints, and real lessons learned, not surface-level overviews. Talks range from architecture and model behavior to tooling, workflows, and experiments that worked or failed. It’s a place to see how people are pushing beyond simple demos into usable systems. Ask questions, and connect with others working in the space.
Immediately after the talk, we will be hosting Open Hack Night, so stick around to meet other Hackers Guild members, enjoy free pizza and popcorn, and see what everyone is working on.
Event Venue & Nearby Stays
Hackers Guild, 2247 Babcock Boulevard, Pittsburgh, US
USD 0.00





