Gabe

Voice AI Engineer @PodiumHQ | Builder of HSMs, Agents & AI infra | Obsessed with observability, performance & reliability.

[Gabe @glwillen]

@thesupermanmx This one is better https://t.co/EuNIqMOhBV

[Gabe @glwillen]

Watching a rocket float in the ocean was awesome!

[Gabe @glwillen]

@fdotinc I have this and it's 100% local and multimodal. Over 2 years of work with the goal of helping my father-in-law with dementia. It just happens to also work incredibly well with agents.

https://t.co/NGhBS4ILwr

[Gabe @glwillen]

@morganlinton Check this out!

https://t.co/dT3uxir4YW

[Gabe @glwillen]

OpenClaw users: tired of losing context or paying for weak summaries?

Just dropped Cortext - a living memory engine for OpenClaw.

Every message saved durably to on-device SQLite. Per-conversation isolation (no leaks). Live recall on every turn. Zero LLM calls for compaction. Real 372k token sessions shrink to ~1.5k instantly with better recall than summarizers. Fully offline after first run.

Install: openclaw plugins install @augmem/cortext-openclaw-plugin

This changes how long-running agents actually remember.

https://t.co/GgUPua3gia

#OpenClaw #AgentMemory #LocalAI @steipete @openclaw

[Gabe @glwillen]

openclaw plugins install @augmem/cortext-openclaw-plugin

Try it!

@steipete

[Gabe @glwillen]

@pkursawe @DavidKPiano https://t.co/yorSJQMlz8

[Gabe @glwillen]

@pkursawe @DavidKPiano There is a WIP compiler that will translate between every language and other frameworks including xstate.

[Gabe @glwillen]

@DavidKPiano @pkursawe Both are compile time libraries and add very little runtime overhead.

[Gabe @glwillen]

@DavidKPiano @pkursawe I have a couple
https://t.co/9wvnypNfHU -> SML with fixes
https://t.co/BLiAMZJAAD -> Easier DSL but slower compile times

[Gabe @glwillen]

@brettcalhounn 🙋‍♂️

[Gabe @glwillen]

@simonw Testing

[Gabe @glwillen]

@hthieblot Except death

[Gabe @glwillen]

@thsottiaux I sent you a DM!

[Gabe @glwillen]

Are we still talking graphs or did we shift to quantum yet?

[Peter Steinberger 🦞 @steipete]

Are we still talking loops or did we shift to graphs yet?

[Gabe @glwillen]

@thdxr I'm gonna need you to go back to being a sarcastic smart ass.

[Gabe @glwillen]

@GoogleAIStudio A better version of Gemini...

[Gabe @glwillen]

@DavidKPiano Can I thank you now? Otherwise i'll forget.

[Gabe @glwillen]

@grok + @warpdotdev = 🔥

[Gabe @glwillen]

If capacity limits require interrupting agent threads, that's understandable. But it would be great if interrupted threads automatically resumed once capacity became available again. @thsottiaux