Cover Image for Essence VC Q2 Virtual Conference: Multimodal AI
Cover Image for Essence VC Q2 Virtual Conference: Multimodal AI
274 Going

Essence VC Q2 Virtual Conference: Multimodal AI

Hosted by Essence Venture Capital & Timothy Chen
Zoom
Registration
Past Event
Welcome! To join the event, please register below.
About Event

​Meet the technical founders and developers at the cutting edge of multimodal AI.

Join the minds behind AI Town, ComfyUI, Genmo, Nexa AI, Rosebud AI and Eventual for an inside look into their latest developments.

Tim Chen of Essence Venture Capital will also host an open panel with our speakers to answer audience questions and share predictions on the future of beyond text LLMs.

Full Conference Agenda (PST):

10:00 - 10:05 → Welcome (Naomi Walker-Garrett & Tim)

Meet the team at Essence Venture Capital: The early stage dev tools and data infra specialist fund.

10:05 - 10:25 → NexaAI, presented by Zack (Zhiyuan) Li

Zack Li is the CTO and Cofounder of Nexa AI which develops energy-efficient and multimodal lightweight LLMs for building AI Agents both on device and cloud. Nexa AI's Octopus-V2 model boasts significant performance improvements over RAG plus open-sourced LLM solution, achieving 35 times faster inference speed and 70 times more energy efficiency. Previously, Zack worked in on-device ML at Google and Amazon Lab126.

In this session Zack will cover Nexa's Octopus models, Coral SDK, Octoverse product and how to join their team.

10:25 - 10:45 → AI Town, presented by Ian Macartney

Ian Macartney is a software engineer at Convex who developed AI Town in collaboration with a16z. He has previously created backend infrastructure for file previews at Dropbox, an ML workflow system for journalists at The New York Times, and other projects across iOS, web, and hardware platforms.

In this session he will discuss the architectural patterns behind AI Town and share insights on designing real-time multiplayer AI applications.

10:45 - 11:05 → Genmo, presented by Paras Jain

Paras is the co-founder and CEO of Genmo, a generative AI startup on a mission to bring video creation to everyone. Genmo is developing leading AI models for high-quality video generation from text. Prior to Genmo, Paras earned his doctorate degree in computer science from UC Berkeley where he developed foundational techniques to scale training of LLMs.

In this talk, Paras will discuss key challenges and innovations towards high fidelity video generation from text.

11:05 - 11:25 → Rosebud AI, presented by Loic Le Feuvre

Loic is Head of Engineering at Rosebud AI, the startup building the AI Roblox. They make it dramatically easier to craft the game of your dreams by going from text description to code to game.

Loic will share what they've built and more importantly how they built it, focusing on what it takes to harness the power of LLMs and make it available to Rosebud’s users (evaluations, tests and experiments, prompt engineering).

11:25 - 11:45 → Eventual (Daft), presented by Jay Chia

Jay is the cofounder of Eventual, the all-in-one platform for modern data workloads across data engineering, machine learning and analytics.

Prior to Eventual, Jay was a software engineer building large scale ML data systems for computational biology at Freenome and self-driving cars at Lyft.

In this session Jay will share what Eventual has been working on with regard to multimodal/AI data warehousing.

11:45 - 12:05 → ComfyUI, presented by ComfyAnon

Comfyanonymous is the author and maintainer of ComfyUI, the most powerful and modular stable diffusion GUI, api and backend with a graph/nodes interface. He's also the cofounder of comfy.org, the team dedicated to advancing and democratizing visual AI access through open source tooling. They work with OSS model builders (see Comfy's recent support for SD3), and organize community efforts to efficiently build OSS tooling for all.

12:05 - 12:25 → Open Q&A Panel (Tim Chen + all speakers)

This is the audience's opportunity to ask our speakers their most burning questions on multimodal LLMs and the future of AI infrastructure.

Led by Tim Chen, Managing Partner at Essence VC, with a decade of experience leading engineering in enterprise infra and open source communities/companies.

Prior to Essence, Tim was the SVP of Engineering at Cosmos, a popular open source blockchain SDK. Prior to Cosmos, Tim cofounded Hyperpilot with Stanford Professor Christos Kozyrakis leveraging decades of research to disrupt the enterprise infrastructure space, which later exited to Cloudera.

12:25 - 12:30 → Outro (Tim & Naomi)

Closing comments and an opportunity for the audience to share their feedback + what they'd like to see from our next conference.

274 Going