Owning the Inference Stack · 01 · with Vidya
- Date
- 2026-05-30
- Host
- Frontier Tower SF
About this event
If you care about how modern AI systems actually run in the real world, this meetup is for you. Owning the Inference Stack · 01 · with Vidya is a focused in-person gathering for people who want to talk seriously about inference infrastructure, practical tradeoffs, and what it means to build with more control over the stack instead of treating deployment as a black box. This first session is also a community moment: a chance to meet others thinking deeply about model serving, performance, reliability, and the operational reality behind AI products. Whether you come with strong technical context or simply real curiosity about how inference gets built and scaled, you should expect thoughtful conversation and a room full of people who care about the details. What Is This? This event is an in-person meetup centered on the idea of owning the inference stack. At a high level, that means looking beyond models alone and paying attention to the systems, tooling, and decisions that determine how inference actually works in practice: speed, cost, observability, orchestration, reliability, and the developer choices that shape product outcomes. The "01" in the title signals the start of a series or a first gathering around this theme. That makes this event especially useful if you want to help shape the tone of the community early, meet others who are interested in the same problems, and be part of a conversation while it is still small enough to feel direct and substantive. With Vidya anchoring the session, the event is designed less like a broad conference and more like a strong meetup: conversational, grounded, and oriented toward real exchange. Expect a format that supports learning from the room, not just listening passively. Because the tags include community, networking, meetup, and social, you can also expect this to be a genuinely human event, not only a technical one. The goal is to bring together people who want sharper conversations about inference while also making space for connection, introductions, and follow-up relationships. What to Expect The evening will likely move through a few distinct modes: framing the topic, surfacing practical perspectives, and creating space for discussion. Rather than trying to cover everything about AI infrastructure, the event is centered on a clear theme, which should make the conversation more useful for attendees who want depth over noise. You should expect a meetup-style flow that may include: A welcome and framing of the topic so everyone has a shared starting point Discussion around inference-stack decisions such as tooling, deployment patterns, performance tradeoffs, and operational ownership Room for questions and perspective-sharing from attendees with different backgrounds Networking time before, during, or after the main discussion Because this is an in-person event, one of the biggest benefits will be the side conversations. Often the most useful insights come from hearing how someone else approached latency, scaling constraints, serving architecture, or internal platform decisions in a real setting. You do not need to show up with a polished point of view. It is enough to come ready to listen closely, ask strong questions, and compare notes with others who are building, exploring, or evaluating inference systems from different angles. Why Attend If you work anywhere near applied AI, the inference layer increasingly matters. Product quality, user experience, operating cost, and system reliability are all shaped by decisions that happen after model selection. This event gives you a place to discuss those choices with people who understand why they matter. You should attend if you want sharper thinking around questions like: What parts of the stack are worth controlling directly? Where do abstractions help, and where do they hide important tradeoffs? How do teams balance speed of iteration with performance and reliability in production? These are the kinds of questions that benefit from live discussion with practitioners and peers. There is also value in the room itself. Good technical communities are hard to find, and even harder to build from scratch. Attending early means you can meet people before the network gets diffuse, have more meaningful conversations, and become part of an emerging circle of people interested in inference as a serious craft rather than a vague trend. You may leave with practical ideas, new contacts, and a clearer vocabulary for discussing inference architecture and operational ownership. Just as important, you may leave with a better sense of who else is working on similar problems and worth staying in touch with. Practical Details This is an in-person event taking place on Friday, May 29 at 6:00 PM PDT. The in-person format matters here: this is the kind of topic that benefits from active discussion, quick back-and-forth, and the informal conversations that happen before and after the main session. Since this is scheduled for the evening, plan for a meetup atmosphere rather than a formal daytime program. That makes it well suited for people coming from work, shifting out of the weekly sprint, and ready to think a little more broadly with others who care about systems, infrastructure, and AI operations. Before attending, it helps to come with a few things in mind: A question you are currently thinking about related to inference, deployment, or model operations A short way to describe what you are working on if you want to make networking easier An openness to both technical and community conversation, since the event sits at the intersection of both If this topic is close to your work or to the direction you want to move in, this is a strong room to be in. Come ready to learn, compare approaches, and meet people who are serious about what happens after the model is chosen.
Who should attend
This event is for people who want a smarter, more grounded conversation about AI inference and the systems behind it. - **You work on applied AI products** and want to better understand the infrastructure decisions that shape latency, reliability, cost, and user experience. - **You are an engineer, builder, or technical operator** who thinks beyond the model itself and cares about deployment, serving, performance, and operational ownership. - **You are exploring the inference layer for the first time** and want a strong entry point through real discussion instead of generic content. - **You like meetup environments where you can both learn and talk**, not just sit through a one-way presentation. - **You want to meet others in the community** who are thinking seriously about AI systems, infrastructure, and the practical tradeoffs behind modern products. - **You value being early to a conversation** and want to help shape a new gathering around a timely, specific topic. If you have been looking for a room where inference is treated as an important product and engineering question, not an afterthought, you will likely feel at home here.