Open Source in Finance Forum London 2026

The FINOS Open Source AI in Finance Leadership Summit is a closed-door, peer-to-peer gathering of the most senior technology and AI leaders across financial institutions. This summit is designed to move beyond high-level theory and focus on actionable, open-source collaboration to solve industry-wide challenges in AI governance, security, and implementation.

Full Event Details

An overview of the Agentic and other AI standards supported by Fluxnova, showcasing how FINOS Fluxnova's deterministic workflow capabilities brings Control, Observability and Audit Traceability to Complex Agentic Workflows.

Financial institutions increasingly operate across multiple service channels including mobile banking applications, payment systems, fraud platforms, contact centres and physical branches. In many organisations these environments remain fragmented, resulting in inconsistent customer journeys and operational complexity.

This session explores the concept of “phydigital banking” - the integration of physical and digital banking channels and why orchestration layers are becoming an important architectural capability for modern financial institutions.

The session will illustrate how process orchestration platforms such as Fluxnova, a FINOS open source project neutrally governed by five major banks, can help coordinate workflows across banking systems and improve operational resilience.

Using practical financial services scenarios, we will demonstrate how orchestration can connect digital banking services with human-assisted channels, enabling more transparent, explainable and efficient operations across the institution.

The session will also touch on the emerging role of agentic process automation, where AI-driven agents coordinate workflows across multiple platforms.

Digital asset platforms operate under demanding conditions: strict security requirements, high availability expectations, and increasing regulatory scrutiny. At the same time, engineering teams must retain the flexibility to evolve their systems as the digital asset ecosystem continues to develop.
This session explores lessons learned from supporting the operation of a regulated digital asset platform built on cloud-native and open source technologies. The talk outlines architectural patterns, operational practices, and governance approaches that enable financial-grade infrastructure while maintaining transparency and control.
This real-life use case highlights what worked for us and what didn't.

In this lightning talk, we'll introduce Symphony's AI Platform: a secure, compliant foundation purpose-built for the financial industry. We will walk through the core design principles that make AI adoption viable in regulated environments, covering how we handle data isolation, access control, audit trails, and model governance at scale.

We'll then show how we extended the platform's capabilities by integrating with Fluxnova. This integration enables financial teams to orchestrate hybrid workflows that combine deterministic, rules-based processes (BPMN) with non-deterministic AI agents.

Attendees will leave with a concrete understanding of what it takes to bring AI into a regulated financial environment, and how open standards like Fluxnova can bridge the gap between structured process automation and modern agentic AI.

Join us for a hands-on look at Fluxnova 3.0 and its latest features for smarter workflow automation: Dive into Fluxnova 3.0 the next evolution of open-source workflow automation! In this session, we’ll unveil powerful new features designed to give you unprecedented visibility and control over your processes.

Financial services is beginning to experiment with multi‑agent AI systems, opening up opportunities for agents to enable and accelerate real operational workflows. This talk introduces Fluxnova’s Workflow Query MCP Plugin as one enabler of that future. By exposing live BPMN state and process metadata through MCP tools, the plugin equips agents with the context needed to interpret and reason about orchestrated FS processes.

This shared visibility unlocks new possibilities: networks of specialised agents that coordinate investigations across settlement, risk, and servicing flows; supervisory agents that explain workflow paths in natural language; and oversight agents that detect emerging friction before it impacts customers.

This talk highlights how richer observability can “connect the dots” between agents, orchestration engines, and human decision‑makers, and attendees will leave with a glimpse of how observability-first design can help shape safer, more collaborative agent ecosystems in financial services.

CALMSuite is a new FINOS project designed to accelerate end user adoption of CALM, focusing on a smooth journey for architects with no previous CALM experience. This presentation will tell the story of how CALMSuite began as a hackathon project in 2026, and quickly became the de facto user experience for users who want to architect secure-by-default infrastructure.

If you’re an architect looking to begin your Architecture as Code journey, or a director looking to make dramatic improvements to your engineering outcomes, this session is the perfect starting point to see how simplicity can drive enterprise-scale transformation.

Financial institutions are modernizing data platforms to support real-time analytics, AI workloads, and increasingly strict security and compliance requirements. Open technologies such as Red Hat Enterprise Linux, OpenShift, and automation frameworks enable organizations to run mission-critical database workloads with greater performance, flexibility, and operational consistency.

In this session, we demonstrate how Microsoft SQL Server 2025 running on Red Hat Enterprise Linux and OpenShift provides a modern, secure, and scalable platform for financial services workloads.

We will showcase automated deployments using the latest Ansible enhancements for SQL Server 2025 on RHEL 10 We will also explore containerized SQL Server deployments on Kubernetes with OpenShift, improvements in SELinux integration and password policy controls, and new AI capabilities in SQL Server 2025 including Retrieval-Augmented Generation (RAG) scenarios.

Finally, we will highlight how Validated Patterns on OpenShift AI simplify GPU management, vector database integration, and GitOps workflows for building enterprise AI platforms with SQL Server as a secure backend for embeddings and analytics.

Financial desktop users run 5 to 15 apps at once. MCP Apps lets AI agents return interactive HTML interfaces inside chat interfaces acting as MCP hosts, but the current model buries these apps in chat history with duplicate copies stacking up. No persistent workspace, no inter-app context sharing, no layout control.

For regulated environments where traders, PMs, and compliance teams need multiple tools visible and auditable simultaneously, this is a potential production blocker. This talk addresses three gaps: app lifecycle (chat widgets into persistent workspace applications), multi-app orchestration (coordinated agentic flows across simultaneous MCP Apps), and layout (giving data-dense financial UIs the screen real estate they need).

The open source ecosystem has the building blocks. Open source layout managers like Dockview and GoldenLayout provide workspace primitives MCP Apps hosts lack. FDC3, maintained by FINOS, adds context sharing and intent resolution across active apps. We combine these in a working prototype, building on our OSFF NYC 2025 session on FDC3 intents as MCP tools.