Open Source Summit Korea 2026

This presentation examines the evolving role of open data in the next phase of artificial intelligence development, and the growing imperative to align innovation with trust. As AI systems become increasingly data-driven, the quality, origin, and governance of data are emerging as foundational issues that will shape not only technological progress, but also public confidence and global competitiveness.
The talk situates these developments within a broader shift toward greater transparency and accountability in AI, reflected in emerging policy directions such as the EU AI Act and Korea’s Korean AI Basic Act. Rather than viewing openness as a risk to be constrained, it explores how responsible data practices and shared governance approaches can enable wider participation in AI development while maintaining integrity and trust.
Ultimately, the presentation argues that the future of AI will depend not only on model performance, but on the strength of the underlying data ecosystem. Building sustainable trust around open data—through transparency, collaboration, and accountability—will be critical to unlocking the full potential of AI at scale.

Large Language Models (LLMs) are widely used to search, summarize, and generate answers. They are powerful, but results are not always stable and can be difficult to verify.

Many systems use Retrieval-Augmented Generation (RAG) to connect LLMs with external data. This helps, but does not fully solve the problem. The same question can produce different answers depending on the model or context, making knowledge hard to manage.

This raises a simple question: where should knowledge be managed?

A similar issue exists in software supply chains. SBOM (Software Bill of Materials) makes components and relationships visible, but is often treated as static data.

Based on hands-on experience designing graph-based knowledge systems, this session introduces a design approach: managing knowledge outside LLMs as structured, persistent data.

We discuss lessons from SBOM and how graph-based approaches—using technologies such as SQL/PGQ or GQL—can help understand relationships and trace changes.

The focus is a practical way of thinking for more open and sustainable knowledge practices.

AI agents are quickly moving from prototypes to production systems that manage APIs, tools, and complex reasoning tasks.

Once these systems are deployed, they often act like black boxes, hiding failures, slowdowns, and unexpected behavior.

This session will show you how to add observability to agentic systems with open-source tools and cloud-native tech.

​Brandon and Mostafa will break down the lifecycle of an AI agent, covering prompt execution, tool invocation, memory access, and multi-agent coordination, and demonstrate how to make each stage observable.

Using a practical, architecture-focused approach, they will show how to:

- Trace agent workflows across distributed systems
- Monitor latency and token usage
- Detect anomalies such as hallucinations, tool misuse, and runaway behavior.
- Connect LLM behavior with infrastructure metrics like GPU, container, and network data.

They will also present a complete reference architecture that uses open-source projects including OpenTelemetry (OTel), Prometheus, Grafana, and some new tools for LLM observability.

Attendees will learn practical ways to build transparent and reliable production AI agent systems.

As LLM-based agents gain access to tools - APIs, databases, file systems, and internal services, the security model changes. The model is no longer only generating text; it is selecting actions and invoking capabilities across systems.

Prompt injection attacks exploit this boundary. between model reasoning & external execution.

This talk examines how tool-enabled agents built on open LLM frameworks expand the attack surface and why traditional input validation approaches are insufficient.

We will analyze concrete failure modes such as:
- Prompt injection vs classical injection: control of model reasoning rather than query structure
- Tool outputs as secondary injection vector in multi-step workflows
- Why system prompts & guardrails are not reliable isolation boundaries
- Capability scoping & least-privilege design for tool access
- Isolation patterns for tool execution (sandboxing, mediated execution layers)
- Structured tool interfaces vs free-form prompting
- Observability patterns for tracing agent decisions and tool calls
- Adversarial testing of agent pipelines

Examples draw from patterns emerging in open-source LLM and agent ecosystems.

In April 2026, x402 joined the Linux Foundation with over 20 founding members. It is the first payment protocol under open source governance at the Linux Foundation.

x402 activates HTTP 402 "Payment Required," reserved in 1997 but never used. A server responds with 402 and payment terms, the client settles on-chain, the resource is delivered. Apache 2.0, zero protocol fees, chain-agnostic, no single-organization dependency.

Part 1: Why now. Every internet layer runs on open protocols except payments. That breaks when AI agents must pay at machine speed. We map the agentic payment stack and ask: is it open, who governs it?

Part 2: How x402 works. HTTP 402 challenge-response, stateless architecture, one middleware line to gate any API, production deployments today.

Part 3: Bottom-up to foundation. x402 was open-sourced, developers adopted it, major infrastructure providers shipped native support, competing protocols chose to integrate x402 rather than build a rival. Then it moved to the Linux Foundation.

Part 4: What comes next. The protocol is early. Foundation governance changes the signal. We examine what sustained adoption requires.

Machine learning is increasingly used in high-energy physics, particularly in trigger systems that process data at rates of 100 kHz while making real-time event selection decisions. The latency and reliability requirements demand highly optimized inference pipelines. While portable solutions such as ONNX Runtime simplify deployment, many applications rely on hardware-specific libraries like NVIDIA TensorRT and MIGraphX (via ROCm) for optimal performance. Code-generation approaches such as SOFIE offer additional efficiency but introduce integration complexity.

We present a unified inference interface that abstracts backend-specific details while preserving performance. It enables execution across multiple inference libraries without data copies or user-side configuration changes. An offline processor converts trained models into backend-optimized plans, and a lightweight runtime loads and executes them through a common API with direct data access for heterogeneous environments.

AI systems need more than intelligence; they need context that persists. Without it, even strong models can misinterpret information, lose decision rationale, or repeat the same mistakes. Context Graphs have emerged as a practical pattern for agentic AI: a living graph that captures not only what was retrieved or known, but how context led to actions through tool calls, constraints, policies, and outcomes, stitched across entities and time so precedent becomes searchable.

This talk explores context engineering as the discipline of designing that context layer, and shows how context graphs complement retrieval by enabling multi-hop, structured context assembly (building on GraphRAG-style hierarchical summaries) while improving explainability and evaluation. Attendees will leave with a practical understanding of how to build context pipelines that combine contextual retrieval with persistent memory and provenance, and why context graphs are becoming central to trustworthy, enterprise-ready AI systems.

As AI agents become increasingly capable of generating code and executing complex workflows, their use in research is still limited by concerns about rigour and reproducibility. This session introduces ado, an open-source framework that brings structure to agent-driven scientific experimentation.

ado defines schemas for the core elements of a discovery process: the problem space and how to explore it. Agents iteratively propose and refine experimental campaigns as validated configurations based on these schemas, while ado handles execution. This separation of research intent from execution constrains agents to focus on the research task, reducing hallucinations and the need to write boilerplate code. Combined with a set of agent skills for formulating problems, creating and running experiments, and analysing their results, ado provides a framework for end-to-end agent-driven discovery workflows.

Whether you are an experienced researcher or new to computational experimentation, this talk presents a practical model for integrating AI agents into research workflows while keeping experimentation structured, transparent, and reproducible.

Standard FP8 formats suffer from frequent gradient overflows and heavy reliance on complex Delayed Scaling, which often lead to training instability or suboptimal convergence in large models. This session introduces HiFloat8 (HiF8) — a tapered precision format that offers an alternative approach to managing dynamic range. This "natural" alignment with neural network weight/gradient distributions allows HiF8 to capture high-magnitude outliers without the aggressive scaling required by standard FP8.We explore how HiF8 can works in the training and inference procedure.We will demonstrate the implementation of HiF8 within the ecosystem. They allow developers to evaluate performance of Hif8 on GPUs. Also, we will give an analysis of training stability and final loss parity where HiF8 provides relatively the same accuracy and 1.5-1.7 times GEMM performance than FP16. Finally, we will share insights from our ongoing collaboration on dedicated hardware support for HiF8.