Events
Rally events, and the rest of AI SF
Builders Night, BSOTA, Speaker Series, and what else is on around the city.
Saturday, August 8
10 AM - 1 PM
Your coding agent resets every session and wastes twenty minutes relearning your repo. This hands-on build sprint teaches you RAG over your own codebase so the agent actually remembers what it figured out yesterday, shipping with persistent memory by 1pm.
2 PM - 5 PM
Haoyi Niu from UC Berkeley walks through ODEWorld, a continuous latent world model that treats prediction as ODE solving in compressed space. If you're building embodied AI, this bridges the gap between planning-friendly dynamics and visual fidelity in ways discrete-time models can't match.
3 PM - 7 PM
As agents move into production, governance becomes non-negotiable. This gathering brings engineers, founders, and policy thinkers to debate runtime controls, audit trails, and compliance frameworks while testing live demos and Tesla's full self-driving system.
6:30 PM - 8:30 PM
9:30 AM - 5 PM
Silicon Valley's biggest robotics gathering brings 5,000+ engineers, investors, and students together for competitions, live demos, and 30+ expert talks across two days. If you're building in robotics or AI, this is where the ecosystem actually convenes.
Tuesday, August 11
5 PM - 7:30 PM
super{set} is hosting an info session on Spark Tank, their founder pitch competition debuting at Rewire in September with a spot at their $200M-backed venture studio on the line. Meet the GP team, ask how participation works, and network with other founders considering the leap.
6 PM - 8 PM
A 5k run along the Embarcadero for founders and founding engineers who need a break from grinding, hosted by Lore and LlamaIndex. All paces welcome, meeting at Sue Bierman Park around 6:15pm for a loop to Pier 39 and back.
7 PM - 12 AM
Slash opens their downtown office twice a month for late-night builder sessions where the only rule is no day job or schoolwork, just personal projects and weird ideas. Bring your laptop, grab free chai and coffee, find your flow with other engineers and designers, and demo what you built at the end of the night.
Wednesday, August 12
1 PM - 7 PM
Boris Cherny and Anthropic's engineering team walk you through building agents at scale, then you ship something real in the workshop using live credits. Bring your laptop and the specific problems bottlenecking your startup right now.
5 PM - 8:30 PM
Sanscritic is exploring how AI systems can evaluate their own work, improve iteratively, and continuously produce better models and products through metareasoning. The evening covers self-improving architectures, simulation-driven development, and the startup opportunities emerging across software, robotics, biotech, and deeptech.
5:30 PM - 7:30 PM
AI coding hype meets reality as builders share what actually works automating the software factory, from GitHub Actions to agent memory techniques. Bring a five-minute demo of a tactical win you've found, skip the pitch, and learn where the real landmines hide.
5:30 PM - 8:30 PM
Hands-on workshop where you'll build a working voice agent using Agora Agent Studio and CLI in a few hours. Solid for developers and founders who want to move from idea to prototype without the usual friction.
6 PM - 7:30 PM
Build a working sales lead agent without touching code in this hands-on workshop. You'll learn how computer-use agents research prospects and automate targeting, with support from the team and a small group of fellow builders.
6 PM - 7:30 PM
Early-stage AI founders get scored on their tech's credibility and defensibility, then top startups pitch directly to VCs and angels. It's a structured way to sharpen your fundraising story if you're actually ready to raise.
6 PM - 8:30 PM
Inference costs and latencies vary wildly across providers, sometimes by 12x end-to-end and 3x in price for the same query. Artificial Analysis benchmarks 500+ endpoints to cut through the noise, with their team and inference leaders breaking down what actually matters for your stack.
7 PM - 10 PM
You'll wire up a swarm of coding agents in parallel across git worktrees so multiple agents tackle independent tasks simultaneously on one repo. Hands-on walkthrough covers the conductor pattern, merge strategy, and when parallelism actually pays off versus tangling your codebase.
Thursday, August 13
12:30 PM - 8 PM
Build AI agents with persistent memory in five hours at MongoDB's .local Build Fest, then demo on stage for $15k and a year of Cursor Ultra. This is where agent memory meets real databases and APIs in a competitive sprint with serious prizes.
4 PM - 7 PM
Scrappy Founders Play Pickleball brings together early-stage founders and investors for competitive games and dinner at private indoor courts. Shows up if you're seed to Series A and want to network with people actually shipping something.
5:30 PM - 7:30 PM
Convex opens their HQ to show how they're rebuilding the backend for modern developers, with the team walking through their tech and mission. Solid if you're evaluating your next startup move or curious how a high-growth infrastructure company actually operates.
5:30 PM - 8 PM
5:30 PM - 8:30 PM
DataFusion is quietly reshaping how companies handle query engines at scale, and this meetup brings together the people actually building with it. Expect real-world use cases and performance deep dives from teams pushing the boundaries of modern data infrastructure.
6 PM - 8 PM
6:30 PM - 8:30 PM
Small group of Series A+ founders sharing unfiltered conversations on fundraising, GTM, and scaling over dinner at a top SF venue. Worth attending if you're at that stage and want tactical insights from peers who've already figured some of it out.
6:30 PM - 9 PM
Watch scientists and engineers demo what they built in a 100-day AI fellowship: open source co-scientists, DIY labs, agentic workflows, and RL systems for gene therapy. Stick around for open mic if you're building something in this space too.
9 AM - 4 PM
MongoDB.local Build Fest brings together AI founders and engineers shipping production agents at Pier 48, with hands-on workshops on LangGraph, vector search, and data patterns from speakers at Rivian and Google. Eligible founders get free access including expert office hours and startup credits through MongoDB for Startups.
Friday, August 14
5 PM - 7 PM
Saturday, August 15
10 AM - 1 PM
This is the finale of a nine-part build series where you assemble a complete coding-agent stack piece by piece, then wire it all together today on hardware you control. By 1pm you walk out with a self-hosted agent running locally, zero API bills, and actually owning the whole thing.
6:30 PM - 8:30 PM
9 AM - 9 PM
Build an autonomous agent that actually runs a business end-to-end, from decisions to operations with minimal human oversight. Terac's backing this one with real resources, so it's worth seeing what actually works at scale.
Sunday, August 16
2 PM - 5 PM
Casual Sunday co-working at Salesforce Park where SF AI builders network, code solo or collaborate, then do a quick standup on what you're working on. All skill levels welcome, whether you're here for coffee, projects, or community.
Monday, August 17
4 PM - 6:30 PM
Nicholas Roberts presents his COLM 2026 paper on Train-to-Test scaling laws, showing how test-time compute changes optimal pretraining decisions. Practical insights for anyone thinking about inference budgets and the overtraining regime that standard scaling laws miss.
5 PM - 7 PM
Neuromorphic chips like Intel's Loihi could cut AI hardware energy use by 10-100x by mimicking how brains actually work. Ramashish Gaurav shares hands-on experience building and deploying energy-efficient spiking neural networks that challenge the assumption that general intelligence requires massive power.
5 PM - 8 PM
Three hours to advance what you're already building, with Workato technical support and structured introductions to potential collaborators. You'll get feedback from serious builders, demo your progress, and leave with clear next steps.
6 PM - 8:30 PM
An hour of 5x5 soccer with AI founders and investors who've actually shipped something, followed by drinks. Best way to scope out the scene if you'd rather run than sit through another pitch dinner.
6 PM - 9 PM
Intimate gathering where founders, engineers, and AI builders swap real stories about shipping, failing, and solving hard problems instead of exchanging LinkedIn requests. The Builder Bingo challenge sparks actual conversations about the work behind the work.
9:30 AM - 7 PM
A full day of talks on agent feedback loops, identity security, and MLflow from engineers at Databricks, CopilotKit, and Descope. Tailored for mid-career builders shipping AI systems, with demo day at the end.
Tuesday, August 18
5 PM - 8 PM
Salt Bread Maxxing brings the community together over fresh shio pan from Tano SF and quick lightning demos of new PromptQL work. It's part casual bakery obsession, part way to see what other Bay Area builders have shipped this month.
5 PM - 8:30 PM
Sanscritic explores AI systems that evaluate their own work, improve iteratively, and continuously produce better models and products through metareasoning. Learn how simulation and recursive self-improvement unlock new startup opportunities across enterprise software, robotics, biotech, and deeptech.
7:30 AM - 7 PM
Modular's annual conference brings 300+ engineers building AI infrastructure to hear product launches and deep dives from teams like Google DeepMind and Qualcomm. Hands-on Mojo workshops and a robotic demo worth seeing justify the day in August.
9 AM - 11 AM
This is a deliberately small breakfast for seed-stage founders who want peer conversations over speakers and structured seating that forces real connections. If you're past the pitch phase and tired of surface-level networking, this one's designed for the founders actually building something.
Wednesday, August 19
4 PM - 6:30 PM
Parth Asawa presents research that benchmarks how frontier AI systems actually adapt across real-world environments rather than solving isolated tasks. Learn why simple context memory outperforms dedicated systems and why even top models capture just a quarter of possible learning gains.
5 PM - 8:30 PM
Sanscritic explores AI systems that evaluate their own work, learn from feedback, and recursively improve models, datasets, and products through metareasoning. The conversation spans new startup opportunities in enterprise software, robotics, biotech, and deeptech enabled by self-improving AI.
5:30 PM - 7:30 PM
Most agent teams nail the demo then hit the wall at scale, production, real edge cases. This is demos and networking with people who've actually solved the demo-to-production gap.
5:30 PM - 8 PM
Enterprise buyers increasingly demand software deployed in their own environments, forcing vendors to rethink deployment models entirely. This panel covers the full BYOC journey: launching self-hosted options, securing access, and operating across dozens of customer deployments without traditional overhead.
6 PM - 8:30 PM
Bay Area Rust folks are back at Convex in August with talks on open enums, lightning demos, and the usual crew debating ownership semantics over pizza. Good for meeting locals actually shipping in Rust and figuring out which debates matter vs which ones don't.
6:30 PM - 8:30 PM
Thirty hardware founders, engineers, and investors breaking bread at Studio 45 to swap asks, offers, and actual connections over three courses. Small enough to matter, intimate enough to build real relationships in the SF hardware scene.
6:30 PM - 8:30 PM
Intimate dinner with Series A+ founders scaling real companies, hosted at a top SF venue. Tactical shop talk on fundraising, GTM, and scaling from peers at your stage.
Thursday, August 20
3 PM - 8:30 PM
Infrastructure builders and platform engineers designing compute, tooling, and deployment systems will hear from industry leaders on scaling AI sustainably while hands-on workshops let you tackle real constraints. The mix of keynotes and live demos makes this a solid entry point if you're thinking seriously about the infrastructure layer.
4 PM - 8 PM
Manufacturing execs and AI founders actually talk candidly over pickleball and dinner instead of cold emails and doomed pilots. Twelve vetted people debating what makes enterprise deals stick or why they die.
5 PM - 7 PM
Finance leaders and founders from high-growth companies gathering on a rooftop to actually talk shop, no panels or pitches required. This is where the people keeping companies running meet each other with views of the city as backdrop.
5 PM - 9 PM
Eight companies ship live product demos as the sun sets over the city, then you network with the builders behind them. This is the rare mixer where you actually see what people are working on instead of just exchanging cards.
5:30 PM - 7:30 PM
A meetup for people selling GPUs, hardware, storage, and cloud to AI builders, born from GTC conversations. Useful for comparing notes on infrastructure trends and meeting others in the same sales space.
6 PM - 8 PM
Parag Agrawal on why web search needs to be orders of magnitude cheaper as agents take over, with technical specifics on building agents at scale. A rare interview with the Parallel founder and former Twitter CEO who's thinking past chatbots.
6 PM - 9 PM
Six hardware founders pitch frontier physical tech to a live VC panel dishing real-time feedback. This is the kind of in-person deal-flow event that only works when hardware builders, investors, and operators actually fill the room.
6 PM - 9 PM
Hands-on night of building a dataset and fine-tuning an open model on real GPU hardware without worrying about costs. Walk out with a working tuned model and practical experience in the actual workflow.
7 PM - 10 PM
A casual night with founders and engineers actually shipping agent systems, playing pool between conversations about production bottlenecks and new architectures. Competitive spirit required, pool skill optional, dinner and drinks included.
8 AM - 9:30 AM
Casual breakfast gathering for people actively building AI agents inside tools like Claude, Cursor, and OpenClaw. No agenda or pitches, just conversations and demos with founders and PMs actually doing this work.
Friday, August 21
6 PM - 10:30 PM
Vercel and Mantis carved out a private suite at Oracle Park for Noah Kahan on August 21, mixing live music with founder conversation. If you're building something fast-growing, this is the scene to catch the room without the typical networking event energy.
6 PM - 9:30 PM
Three hours to turn an idea into a real project using AI coding agents, no coding skills required. Builder Pub's beginner-focused workshop wraps with feedback from experienced engineers and a community presentation of what everyone created.
Saturday, August 22
6 PM - 9 PM
Hands-on fine-tuning on real Nebius GPUs in three hours, starting from scratch with your own dataset. Walk out with a tuned model and no cloud bill surprise.
6:30 PM - 8:30 PM
Monday, August 24
5 PM - 8 PM
Learn how fast-growing SaaS and AI startups actually build GTM stacks that scale without operational chaos. Covers AI-powered prospecting, essential tech, and connecting data across product and sales.
Tuesday, August 25
5 PM - 6 PM
Weekly breakfast for people shipping production agents, no panels or pitches, just hallway conversations with founders and engineers three months ahead of you. Invite-only seats with YC founders and people from Meta, NVIDIA, OpenAI, and Walmart who are solving the same scaling problems you're about to hit.
5 PM - 8 PM
Most enterprises are running 3-10x more AI agents than they realize, creating a sprawling attack surface that's hard to govern. This session covers discovering shadow agents, securing their actions on enterprise data, and responsibly scaling AI adoption with live demos from the teams building observability layers for production agentic systems.
6:30 PM - 8:30 PM
Wednesday, August 26
5:30 PM - 8 PM
Daytona's exploring runtime infrastructure for long-running agents, with a confirmed talk on forkable sandboxes that let you pause, branch, and sample continuations from machine state. Worth attending if you're building systems where agents need persistent state across sessions.
8:30 AM - 5 PM
TailscaleUp brings engineering and security leaders together to discuss secure connectivity at scale, with speakers from GitHub, AWS, and Grafana Labs. The agenda covers infrastructure, services, and device networking across modern teams.
Thursday, August 27
6 PM - 8 PM
Hiring globally for AI teams means navigating visa complexity, timezone coordination, and compliance at speed. This panel covers what actually works from leaders in recruiting, HR tech, and legal.
Friday, August 28
6:30 PM - 9:30 PM
Founders hitting the $500K-$10M ARR inflection point will hear from operators who've scaled beyond themselves: real channel selection, budget ranges, and pricing strategy when margins are tight and switching costs are low. This is the follow-up to a packed first session, built on specifics rather than playbook platitudes.
9 AM - 7 PM
A one-day hackathon in Japantown tackling real problems: reading crises, loneliness, climate action, and beyond. Build with ClickHouse and Postgres, get help from partner engineers all day, and compete for cash prizes and cloud credits.
Past events
A weekly co-working session where builders show up with projects and leave with momentum through focused deep work and supportive share-backs. Perfect if building alone feels isolating and you need the energy of working alongside people actually making things.
Weekly coworking for builders across software, hardware, business, and creative projects. Two hours of focused work alongside others, plus a share-back where you learn what stuck people and what shipped.
Twice-monthly working sessions where builders using Claude Code, Cursor, and custom agents show real setups and workflows, then spend the night comparing approaches and solving problems together. This is where isolated experimentation becomes shared knowledge.
Anthropic researchers found evidence that language models might have something like conscious access, using global workspace theory as their lens. David Valdman walks through the interpretability paper and what it reveals about how information flows inside LLMs.
A weekly coworking session built for side-by-side building, not networking. You'll show up with a project, sprint in focused silence, then share what you made or got stuck on with people who get it.
Most builders using Claude Code, Cursor, or custom agents are still figuring out workflows alone—BSOTA gets you in a room with others doing the same work. You'll see live walkthroughs of real setups and spend the rest of the night comparing notes and debugging approaches with people who actually use these tools daily. Bring your laptop and a project you're stuck on.
A weekly coworking session where builders show up with projects and leave with real momentum through focused work and peer support. The quiet deep work sprint is interrupted only by genuine moments of discovery, making it one of the few spaces where building alongside others actually feels less lonely.
A weekly coworking session built for the lonely part of building, where you show up with a project and leave with momentum from working alongside others. Quiet deep work sprints with a share-back to learn from what everyone's making.
Most builders using Claude Code, Cursor, or custom agents are figuring out workflows in isolation—BSOTA fixes that. You'll see live walkthroughs of real setups from active users, then spend the rest of the night working alongside other builders who are deep in the same problems. Bring your laptop and a project you're stuck on.
A weekly coworking session where builders show up with projects and work side by side through focused sprints and share-backs. The antidote to solo building: momentum, accountability, and the particular energy of making things real alongside others.
Every Monday, builders bring a project and work in focused silence alongside others doing the same. You'll get 90 minutes of deep work time, then a structured share-back where people show progress, get unstuck, and offer real help. It's accountability without pressure — just the momentum that comes from building in the room with people who get it.
Most builders using Claude Code, Cursor, or custom agents are still figuring out workflows alone—this session gets you in a room with others doing the same work. You'll see live walkthroughs of actual setups and spend the rest of the night comparing notes with people solving similar problems. Designed for regular users who want to level up by learning what's actually working for other builders.
Most networking events end when people leave. Keep in Touch adds structure to what breaks: matching you with relevant people upfront, then using Signal the next morning to actually follow up.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Bi-weekly community event for builders working with agentic coding tools, with techniques going back to a shared repo at github.com/bsota/collective. RSVPs are open.
The first wave of AI rebuilt the digital 15% of the economy. The next wave runs on the other 85%: manufacturing, energy, infrastructure. Atharva Atre (Niantic Spatial) walks through the five technical layers it runs on, from simulated training environments to spatial intelligence.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Builders Night
Monday · 2026-05-11 · 6:30 PM - 9:00 PMMonday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Most AI benchmarks ask whether a model is smart. HumaneBench asks whether it's actually good for the person using it. The team tested how models hold up when users are stressed or vulnerable, and found that 67% of the time the model stops looking out for the user. Talk on the methodology, then a working session on what a benchmark like this would look like for your own product. Best for builders who want to ship AI that doesn't quietly fail the users who need it most.
Spend a Saturday going from tokenization to training a working language model from scratch. You'll write the core architecture, understand the math behind attention mechanisms, and leave with code you actually understand. Best for engineers who want to stop treating LLMs as black boxes.
A deep dive into the design decisions that shape agent architectures, from tool use patterns to memory systems to orchestration strategies.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
A Rally SF hackathon. Why can't the most technical community on Earth coordinate dinner?
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
BSOTA #2. Hands-on session exploring agentic workflows and collaborative AI building.
This session explores AI alignment through Anthropic's Constitutional AI framework. We'll dig into the 84-page constitution that guides Claude's behavior, the difference between rules and dispositions, and how Constitutional AI compares to RLHF. Co-facilitated by Anup Gosavi and Emily Hough-Kovacs. Pre-read: "Constitutional AI: Harmlessness from AI Feedback" (Bai et al., 2022).
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Monday co-working for AI builders. Two hours of focused build time with a share-back round to close.
Justin Lebar (ex-NVIDIA, Google TPU) broke down GPU architecture fundamentals and what they mean for the economics of AI inference. Standing room only.
Weekly co-working session for AI builders. Laptops out, projects in progress, peers to get unstuck with.
Our inaugural BSOTA session. Everett Kleven led the group through shared vocabulary, live demos, and breakout discussions on agentic computing workflows with Claude Code. 93 RSVPs, 40 attended.
We surveyed practical approaches to long-context LLMs including RoPE scaling, sparse attention, and state-space alternatives, then mapped the real engineering tradeoffs: when expanding context windows makes sense versus when retrieval-augmented generation still wins. The takeaway: accepting tokens isn't the same as using them well.
We examined how linear directions in a model's activation space correspond to specific character traits like sycophancy, hallucination propensity, and moral reasoning. The group discussed what personality means for language models and the implications for alignment, safety, and developer tools. Paper: "Persona Vectors" (Chen et al., 2025).
Our first session explored the foundational paper behind modern generative AI. We broke down how diffusion models gradually add noise to data and learn to reverse the process, why this approach outperformed GANs and VAEs, and the connections to Latent Diffusion and Consistency Models. Paper: "Denoising Diffusion Probabilistic Models" (Ho et al., 2020).
Our inaugural session explored DeepSeek-R1 — how reinforcement learning for reasoning, GRPO, and distillation are reshaping what open-source models can do.
Don't miss the next one.
Get the weekly Sunday shortlist of what's worth your time.