Building AI at Scale

AI DevSummit returns to South San Francisco on May 27–28, 2026, bringing together engineers and technical teams responsible for turning AI research into working software. The event focuses on the people who build and deploy real AI systems - AI engineers, software developers, data scientists, and technical leaders working on production environments.

Rather than speculating about the future of AI, the conference centers on how teams actually build, run, and maintain machine learning systems today.

Many conversations around AI still revolve around models and research breakthroughs. But once a model moves into production, the focus shifts quickly toward engineering questions.

Teams have to design training pipelines, manage inference performance, monitor systems, and integrate models into larger products. AI DevSummit reflects that shift, with sessions covering machine learning engineering, enterprise AI deployment, open-source tooling, conversational AI, and neural networks.

The emphasis tends to fall on the practical side of AI development - what happens after a model leaves the research stage and becomes part of a real system.

Behind every AI application sits a larger stack of infrastructure and workflows. Engineers attending the summit are often responsible for building those layers: data pipelines, training environments, model deployment systems, and the platforms that support them.

Discussions frequently focus on issues such as:

• scaling inference workloads

• optimizing training workflows

• integrating AI into existing products

• managing reliability and observability for AI services

These challenges are less visible than model development itself, but they often determine whether an AI project succeeds in production.

AI DevSummit runs as a two-day conference with technical sessions, an expo floor, and a hackathon where teams build and present projects during the event.

That format tends to attract developers and engineers actively working on AI systems. Instead of purely theoretical discussions, many conversations revolve around implementation details - what tools teams are using, how their systems are structured, and what problems they encountered along the way.

As AI continues moving from experimentation into real products, engineering challenges become more visible. Events like AI DevSummit bring together the people solving those problems - the teams designing machine learning pipelines, integrating models into platforms, and operating the infrastructure behind AI applications.

In South San Francisco this May, the most valuable insights will likely come from engineers sharing how their systems work once the models are running in the real world.