AI Infra Signal: Build, Operate, and Monetize Intelligence
The whole stack is still moving — the models, the systems that train them, the systems that run them in production. None of it has settled yet.
Most AI events optimize for volume. This one optimizes for depth — a room with the size enough that every conversation starts already past the introductions, built around the open questions speakers are still working through, not the ones they've solved.
🎤 Speaker Lineup
Keynote Speakers
Nilesh Shah, VP @ hosted.ai
Edward Huang, Partner @ Llama Ventures
The AI Infrastructure Investment Insights
Panel Speakers
Vandon Duong — Principal @ DCVC
Yihua Cheng — CTO @ Tensormesh
Rebecca Biju — AI Infra Engineer @ Together AI
Sachin Keswani — Senior PM @ Nutanix
Lightning Talks
Section 1 - Researchers
Tianyi Zhang — AI Researcher @ Together AI
Optimizing LLM Inference Through Training: Faster Speculative Decoding and Better Low-Bit Quantization
Harshit Joshi — CS Phd @ Stanford SAIL
Contexts are never long enough
Yuexing Hao — Researcher @ Microsoft & Postdoc @ MIT & Founder of MatrAIx AI & Xiaomin Li — Senior Research Scientist @ Microsoft
MatrAIx Simulating the World with 8.3 Billion Persona Agents
Yash More — Member of Technical Staff @ Cohere
Section 2 - Industries
Biwei Huang — Founder @ Aether AI & Assistant Professor @ UC San Diego
Causal world models for the next AI paradigm
Homer Wang — Head of Product @ TinyFish
Dexter Horthy — Founder @HumanLayer (YCF24)
Software Factory Design Patterns
Angelina You — Co-Founder @ Modo (StartX F25)
Methods to evaluate the workflow's ROI and project monetization.
What's Being Debated Across the Stack
Compute, training infrastructure
Inference, agent infrastructure
Reliability, evals, and recursive self improvements
Data, security, and enterprise deployment
Frontier research shaping what comes next
Why join us:
Three keynote-length talks + Eight lightning talks more move fast across two tracks: AI Data / Infra + Researcher Topics
Every talk ends with something unresolved. Those open questions carry straight into the salon
100 people, application-only: founders, frontier researchers, enterprise infrastructure leaders, and infra-focused investors
Format:
Panel Discussion — The future of AI Infra, Agent token-economics and models
Keynote Block — 3 companies, extended time to go deep
Lightning Talks I — AI Infra & Data (4 companies)
Lightning Talks II — Research Topics (4 companies)
Technical Salon — researchers, founders, enterprise leaders and investors on what's structurally becoming the bottleneck
Schedule:
2:00 PM — Doors open, coffee & booth preview
3:00 PM — Opening
3:10 PM — Technical panel
3:40 PM — Keynote block
4:20 PM — Lightning Talks I: AI Infrastructure & Data
5:00 PM — Lightning Talks II: Research
5:45–7:00 PM — Networking, food & drinks
Who should apply:
Founders and CTOs building infrastructure, researchers from frontier labs, enterprise leaders deploying AI at scale, and investors focused on infrastructure. This isn't for general AI enthusiasts — it's for people building or deciding at the infrastructure layer.
🔬 Researchers / Technical Builders -= Frontier lab researchers, PhDs, postdocs, and engineers working on AI systems, distributed training, inference, and agent infrastructure.
🤝 Strategic Partners - Sponsors Infrastructure providers, developer platforms, compute and cloud partners, and ecosystem partners looking to connect with a curated pipeline of AI infrastructure founders.
🚀 Founders Startup Teams - AI infrastructure, agent systems, inference & serving, training infrastructure, data & security, and frontier AI application startups.
📈 Investors / VCs - Pre-seed to Series B infrastructure investors, technical VCs, and angels looking for early access to high-quality AI infrastructure companies.
100 spots. Applications reviewed on a rolling basis.
Why AI Infra Signal
AI infrastructure is moving from research prototypes into production systems — real deployments, real reliability constraints, real scale problems. This gathering brings together the founders, researchers, investors, and strategic partners actively shaping that transition: from lab benchmark to production system, from experimental architecture to infrastructure enterprises depend on. Reach out to the OpenStages team at partnership@openstagesai.com.
Join as Strategic Partners
Community Partner · Enterprise · Startups · Investors
Strategic partners get direct access to a curated pipeline of AI infrastructure founders, the researchers and technical builders shaping the space, and structured follow-up after the event. Reach out to the OpenStages team at partnership@openstagesai.com.
About the host:
OpenStages is a Bay Area team that plans and hosts curated events for founders and builders: hackathons, salons, and closed-door sessions across AI and tech. LinkedIn · openstagesai.com
About Sponsors:
Llama Ventures
We are where insight and AI shape the future.
Llama Ventures is a Silicon Valley–based early-stage venture firm investing in AI companies from Pre-Seed through Series A. With offices in Sunnyvale and San Francisco, the firm has invested in 60+ companies across AI infrastructure, enterprise software, healthcare, fintech, and deep-tech, and manages $300M+ in direct fund capital alongside a $600M+ fund of funds.
Founded by entrepreneurs and technology leaders, Llama Ventures partners closely with founders, providing strategic support and access to a global network of operators, engineers, customers, and investors.
Contact us: info@llamaventures.vc
Hosted.ai
hosted·ai is the operating system for AI infrastructure. The company's GPUaaS software platform transforms the economics of GPU cloud for service providers, neoclouds, and regional infrastructure operators, enabling them to launch, manage and profitably scale GPU-as-a-Service without prohibitive hardware CAPEX. Through GPU pooling, optimized multi-tenant workload placement, and configurable GPU overcommit (2x–10x), hosted·ai delivers up to 5x improvement in GPU utilization, 5x reduction in CAPEX requirement, and 5x boost in profitability versus traditional GPU passthrough models.
The platform includes a rebrandable self-service customer portal, built-in billing integration, and a ready-to-go GPU marketplace, giving operators everything needed to go from deployment to revenue-generating neocloud.