Hack Day: llama.cpp
Welcome to AI LA’s Inaugural Hack Day
We’re kicking off AI LA’s first-ever Hack Day — a developer-first event designed to go beyond surface-level demos and dive deep into the technical foundations of modern AI.
This day is built for builders: the engineers, tinkerers, and open-source enthusiasts who want to understand how things actually work under the hood.
The focus of this inaugural edition is Local LLMs — large language models that run entirely on your own hardware, powered by llama.cpp.
Event Flow
Morning Workshop
Guided setup of llama.cpp on participants’ devices
Deep dive into quantization, embeddings, and performance tuning
Integrating local inference into applications
Best practices for running models efficiently on different hardware setups
Afternoon Hackathon
Teams form to prototype local-first AI tools
Build private assistants, summarizers, mobile AI utilities, or agentic workflows
Showcase your project and get feedback from peers and judges
What Is llama.cpp?
llama.cpp is an open-source C/C++ library that makes it possible to run modern large language models — like Meta’s Llama 3, Mistral, or Gemma — directly on your laptop, desktop, or even edge devices.
It’s the engine behind tools like Ollama, LM Studio, and koboldcpp, enabling efficient CPU and GPU inference through quantization and clever memory optimizations. In short: it’s what brings AI offline.
With llama.cpp, developers can:
Run powerful models locally — no API keys, no cloud dependencies
Fine-tune and quantize models for small devices
Embed LLMs directly into desktop, mobile, or IoT applications
Keep data private while maintaining top-tier performance
This event is about learning how to actually use it.
The Experience
Hosted at AI LA’s Venice HQ — this hack day blends hands-on learning with community. You’ll collaborate with other developers, explore cutting-edge open-source tools, and see how local inference is redefining what’s possible at the edge.
Whether you’re an AI researcher, engineer, or indie hacker, you’ll walk away with real code, real insights, and new connections.
Meet Tejas Rajurkar (Workshop Host)
Tejas Rajurkar is an AI Engineer specializing in real-time animation and emotionally responsive AI agents powered by LLMs and computer vision. At AMGI Studios, he leads initiatives integrating generative AI with interactive media to enhance human-computer interaction. He’s passionate about open-source innovation and exploring how on-device LLMs can bring creative AI experiences to life.
About AI LA
AI LA Community, Inc. is a 501(c)(3) non-profit that's been building community around the responsible development and application of AI since 2016. We host over 100 events a year, including AI on the Lot, Earth Summit, Life Summit, Responsible AI Symposium, and monthly community meetups across various sectors.
About AMGI Studios
AMGI Studios is an independent animation and gaming company that blends cutting-edge AI with storytelling and interactive entertainment. They focus on creating lifelike, character-driven experiences that respond to language, emotion, and visuals—bringing games and virtual worlds to life in new, deeply personal ways.
Interested in sponsoring?
Your support helps keep the LA developer community well-fed, caffeinated, and inspired.
Contact: Fil Graniczny at [email protected]
Members & Sponsors
Become an AI LA+ member to gain access to exclusive workshops and meetups and save over $400 in event prices!
Interested in sponsoring or providing free products to our vibrant community? Reach out to us at [email protected].
Terms and Conditions and Privacy Policy
By registering to an AI LA Community, Inc event, you agree to our Terms and Conditions and Privacy Policy.
Disability Notice
Individuals with disabilities who need accommodations to attend this event should contact [email protected] with their name and contact information. We request that individuals requiring accommodations notify us at least 7 days prior to the event. Every reasonable effort will be made to provide reasonable accommodations in an effective and timely manner.