Tech Week.San Francisco.

Bria × fal × LTX: What's New in Open-Weights Pipelines

Bria + Fal + LTX

A breakfast with Bria, fal and LTX covering open-weight generative media pipelines, followed by Q&A and time to build.

Why go: Three short technical talks plus open building time give this a concrete focus: what to assemble, control and ship in a media pipeline.

People: Guy Bar Hava (LTX) · Gal Davidi (Bria) · Lovis Odin (fal)

Bring pipeline questions. Breakfast starts at 8:30 AM; talks at 9:15 AM.

Fireworks x a16z: Official Tech Week Kickoff w/ Vercel, ElevenLabs, & Stripe

Fireworks + Stripe + Vercel

Fireworks and a16z host a rooftop kickoff supported by Vercel, ElevenLabs and Stripe, with music and networking.

Why go: A broad AI founder and operator gathering to start the week. Choose it for conversations and new connections, rather than a technical agenda.

Exact address is shared after RSVP.

Google for Startups: Engineering 10x with Dr. Astro Teller and Google DeepMind

Google for Startups

Google for Startups brings Astro Teller and Google DeepMind speakers together for sessions on ambitious startup building and physical AI.

Why go: A named speaker lineup pairs moonshot thinking with product and engineering questions. The expert lounge is a useful complement to the main-stage talks.

People: Astro Teller (X) · Paige Bailey (Google DeepMind) · Andrew Mauboussin (Surge AI)

Acceptance applies to this time slot.

Claude Founder House

Anthropic

Coworking, technical office hours and a three-day program of agent, robotics and startup sessions at Anthropic's Founder House.

Why go: Bookable time with the technical team makes this especially useful if you are already building on Claude. Choose individual sessions; this is not a continuous three-day program.

People: Ina Chun (Anthropic) · Caitlin Kalinowski (Anthropic) · Gagan Bhat (Anthropic)

Daily sessions; check the organizer agenda for opening hours.

Google for Startups: From Frontier to Production with DeepMind & Wiz: Building Secure, Scalable AI

Google for Startups

Google for Startups hosts an afternoon of security, agent engineering, unit economics and demos with Google DeepMind and Wiz.

Why go: A broad but practical program for teams moving beyond proofs of concept, with specialist sessions and opportunities to ask architecture questions.

People: Jiong Liu (Wiz) · Jaclyn Konzelmann (Google Cloud) · Matt Ridenour (Google DeepMind)

Acceptance applies to this time slot; individual sessions have their own agenda.

The Future of Inference | General Compute x Fenwick

Fenwick & West

Fenwick and General Compute host a panel about competing inference architectures and deployment at scale.

Why go: Representatives of several compute approaches share one panel. Useful for comparing infrastructure tradeoffs, with time for questions and networking.

People: Jason Goodison (General Compute) · Gill Verdon (Extropic) · Aaron Melgar (Fireworks AI) · Anton McGonnell (SambaNova)

State of Models by OpenRouter

OpenRouter

OpenRouter examines real-world model usage, cost and quality tradeoffs in a discussion with Artificial Analysis and CapitalG.

Why go: A particularly relevant room for readers choosing models: platform usage data and independent analysis offer two different views of what developers actually adopt.

People: Peter Walker (OpenRouter) · George Cameron (Artificial Analysis) · Jane Alexander (CapitalG)

Camp AI: Production-Ready Agents

Auth0

Auth0 brings speakers from Anthropic, Cloudflare and Browserbase together for a program on securing and scaling production agents.

Why go: The lineup spans authentication, browser automation and cloud infrastructure, making the discussion relevant to the hard parts after an agent demo works.

People: Den Delimarsky (Anthropic) · Rita Kozlov (Cloudflare) · Paul Klein IV (Browserbase) · Gareth Davies (Auth0)

AI Heist Challenge with Cogent x Modal x LangChain

Cogent + LangChain + Modal

Cogent, Modal and LangChain host a team-based agent challenge in an isolated, fictional cloud environment.

Why go: A hands-on way to explore agent behavior under constraints. Bring a laptop and use the challenge to test your debugging skills with other builders.

Participate solo or in a pair; bring a laptop.

Google for Startups & NVIDIA: From Compute to Competitive Moats: Building Defensibly.

Google for Startups

A Google for Startups and NVIDIA morning on accelerated infrastructure, inference, agent security and defensible AI businesses.

Why go: The program connects architecture to business decisions, with speakers from NVIDIA, SandboxAQ, Temporal and a16z rather than a generic AI overview.

People: Jen Hoskins (NVIDIA) · Arman Zaribafiyan (SandboxAQ) · Maxim Fateev (Temporal) · Sarah Wang (a16z)

Calendar lists 9:30 AM; the description mentions a 9 AM start. Confirm arrival time with the organizer.

Creator Clubhouse with Replit, Gamma and ElevenLabs

Passionfroot + Figma + Gamma + Replit

Passionfroot hosts a panel and networking with creator partnership leads from Replit, Gamma and ElevenLabs.

Why go: The agenda focuses on how partnerships are selected and sustained. A practical fit for creators building a business around AI and software audiences.

People: Katherine Gilligan · Connie Yao (ElevenLabs) · Fiona Turko (Gamma) · Tala Awwad (Replit) · Jen Phan (Passionfroot)

The calendar has no end time; the organizer agenda estimates an 8:30 PM finish.

Vibe Coding Demo Night with Lovable & Atlassian

Atlassian + Lovable

Lovable and Atlassian host a demo evening for people building with AI development tools.

Why go: A compact way to discover actual projects and talk to their builders. There is a separate application if you want to demonstrate your own work.

Demo applications use the separate form on the organizer page.

AI Tool Evaluations with ScaleKit, SafeMCP, Workato, and AAIF

Workato + ScaleKit

A Workato-hosted evening on AI evaluation, purchasing frameworks and governance, with a keynote and executive panel.

Why go: Useful for teams making a build-versus-buy decision or defining what an agent must pass before it reaches production.

People: Ravi Madabhushi (Scalekit) · Adam Seligman (Workato) · Nnenna Ndukwe (moderator)

Doors open at 5:30 PM; the keynote begins at 6 PM.

3M to 100M Users: How Gamma Actually Did It

AKG Media

Gamma CEO Grant Lee joins AKG Media for a conversation about creator-led distribution, with a preceding growth panel.

Why go: An operational look at distribution with people running creator programs. Useful if your next challenge is adoption rather than another product feature.

People: Grant Lee (Gamma) · Fiona Turko (Gamma) · Eduardo Sanchez-Ubanell (Wispr Flow)

The calendar ends at noon; the agenda offers optional networking until 1 PM.

Aaron Levie + Tanay Kothari @ AGI House | Turing Frontier Fellows

Turing

Two fireside conversations at AGI House with Wispr Flow CEO Tanay Kothari and Box CEO Aaron Levie.

Why go: Two clear product angles in one afternoon: voice interfaces and enterprise workflows. Allow for travel to Hillsborough, outside central San Francisco.

People: Tanay Kothari (Wispr Flow) · Aaron Levie (Box)

Private, curated gathering. RSVP approval is required. Hillsborough is outside San Francisco.

Deployment Is the Moat with Jason Ma (Dyna Robotics)

Savant Ventures

Dyna Robotics co-founder Jason Ma discusses deployment data, reliability and bringing robots into customer environments.

Why go: A sharply defined technical topic and a named practitioner make this a strong choice for readers following physical AI beyond prototype videos.

People: Jason Ma (Dyna Robotics)

Doors at 5:30 PM; the fireside conversation starts at 6:30 PM.

a16z speedrun AI Faire

a16z speedrun

The speedrun AI Faire offers live B2B AI product demos and direct conversations with early-stage founders.

Why go: A useful browse for operators evaluating tools: there is no formal stage program, so you can spend time on products relevant to your team.

Apply for an entry slot and wait for a confirmation before attending.

Talk to Users Hackathon

Stripe + Base44, Cognition, Sense AI + Daytona, FishAudio + MongoDB, BrightData, Clerk

A full-day hackathon pairs technical teams with target users for feedback while they build and iterate.

Why go: The user-testing format is the differentiator. It puts product feedback inside the build sprint instead of waiting for a final judging presentation.

Technical builder event. Check team and application requirements before planning your day.

Computer-Use Research Salon

Quetta Labs + Pinetree Research

Quetta Labs and Solari convene a small salon on computer-use agents, software environments and world models.

Why go: A focused discussion format for researchers and technical founders. The host specifies Chatham House Rule, so attend to exchange ideas rather than report quotations.

Limited seats; conversation follows Chatham House Rule. End time is not listed.

SpeedHacks: The Official speedrun Hackathon

speedrun + speedrun Alpha

A 12-hour speedrun hackathon challenges teams to build a product that removes a meaningful bottleneck in daily work or life.

Why go: A clear problem brief, a full build day and final demos make this a purposeful way to finish SF Tech Week.

Apply with a team or find one at the event. Bring a laptop.