On-Device Agents with Liquid AI

Get ready for a fun, nerdy evening all about on-device AI agents! 🚀

We're bringing together engineers and researchers to explore what it really takes to run powerful AI on tiny devices. We'll kick things off with food, drinks, and some spicy Socratic debates (can a 1B model really compete with a 70B model?! 👀), then dive into technical talks following the journey from Model → Compute → Benchmark. Expect deep technical discussions, big questions, plenty of memes, and zero boring corporate slides. And because the best conversations often happen after the talks, we'll wrap things up at a nearby bar! 🍻

Event Info: https://aisocratic.org/events/on-device-agents

Flow: Model → Compute → Benchmark

Model: Efficient architectures. Open weights. Models designed for local, agentic workloads.

↓

Compute: NPUs, Qualcomm, and Apple silicon. Quantization, memory, and power. Making inference work on-device.

↓

Benchmark: Measure the deployed system. Quality, latency, and efficiency under real constraints.

Speakers

Mathias Lechner · Yeon Kim · Sergii Kryvoblotskyi · Alan Zhu · Henry Ndubuaku

Agenda

• Liquid AI HQ · San Francisco
• Wednesday, November 18, 2026 · 6:00–9:30 PM PST
• https://aisocratic.org/events/on-device-agents

• 6:00 PM · Doors open — Food, drinks, and conversations with the people building the stack.
• 6:30 PM · Meet the room — Host introductions and a quick round of introductions from attendees.
• 7:00 PM · Socratic dialogues — A focused 30-minute discussion on on-device agents: what works, what breaks, and what comes next.
• 7:30 PM · Technical talks + Q&A — Five 15-minute presentations, each followed by 10 minutes of questions.

• 7:30–7:50 PM · Mathias Lechner · Liquid AI — Efficient architectures for local intelligence.
• 7:50–8:10 PM · Yeon Kim · ZETIC — Fast on every chip. Q&A at 8:10 PM.
• 8:10–8:30 PM · Sergii Kryvoblotskyi · MacPaw — From local model to useful app.
• 8:30–8:50 PM · Alan Zhu · Qualcomm — Inference on the NPU.
• 8:50–9:10 PM · Henry Ndubuaku · Cactus Compute — Foundation models in 8–29 MB.
• 9:20 PM · Closing Remarks
• 9:30 PM · Keep the conversation going — Networking at a nearby bar for who's down to join us!

Working titles based on each speaker’s topic; final titles are subject to confirmation. All times Pacific · Program subject to change.

Want to propose a talk or demo? Get in touch.

Guests

For AI researchers and engineers working on small local models and on-device agents. Attendance is by invitation. Request to join on Luma — every request is reviewed to keep the room technical and high signal.

Sponsors

We’re curating sponsors building the on-device agent stack, from models and inference runtimes to hardware, benchmarks, and the apps that ship it. Become a sponsor.

Liquid AI — Presenting Sponsor

Liquid AI mission is to build efficient general-purpose AI at every scale. This means Liquid products are highly capable, compute and cost optimized, and ready to bring intelligence on any device and medium of choice.

AI Socratic — Organizer

AI Socratic brings together engineers, researchers, and founders working on frontier AI. Participation is curated to keep discussions technical, thoughtful, and high-signal. Our mission is to democratize AI with open source and decentralized software, and open dialogues.

Explore Frontier Stack and join the community at aisocratic.org.