October 2025 Summaries

Enterprise-level deployment of agentic AI applications faces challenges beyond technology, necessitating strategic alignment with business objectives to ensure sustainable value rather than remaining in perpetual pilot mode. Key best practices for successful deployment include starting with a clear AI strategy, systematically evaluating potential use cases, and designing AI solutions with production requirements in mind. Establishing comprehensive AI governance, focusing on integration architecture, and prioritizing user experience are crucial elements. Organizations should progressively build internal capabilities, implement continuous improvement processes, and proactively address change management to mitigate resistance. Finally, scaling AI through reusable components can accelerate deployment and ensure consistency, ultimately transitioning agentic AI from experimentation to enterprise-wide application by addressing technology, process, people, and governance comprehensively.

A recent MIT study highlights the "GenAI Divide," where 95% of companies report little to no return on investment from generative AI, despite spending billions. This gap is attributed not to technological shortcomings but to ineffective implementation, with many enterprises using generic AI tools that fail to integrate into complex business workflows. Successful organizations, as identified by the study, adopt customizable platforms, empower business users, and focus on adaptive intelligence that retains context and evolves with user input. Notably, the most substantial ROI is found in back-office operations, where AI optimizes internal processes and reduces external spend. The future of AI in business lies in interconnected, adaptable platforms, and companies must strategically shift towards these systems to remain competitive.

In Q3 2025, Vertesia announced several significant updates and enhancements to its platform, focusing on AI tools, user provisioning, and application development. The company introduced a new framework for building AI applications, featuring a plugin system with a library of UI components, enabling developers to create and scale custom apps efficiently. They launched an Enterprise AI Assistant App, which allows secure interaction with AI agents to perform various business tasks, such as analyzing customer data and generating content. Enhanced provisioning capabilities now support user group management and integration with identity services like Microsoft Entra and Okta, streamlining user access across large organizations. Vertesia also introduced tooling options for AI agents, including tool plugins and an MCP client for accessing third-party tools. Additional features include granular permissions, hybrid search functionality, and enterprise-ready enhancements like Azure AI Foundry support and dynamic rate limiting for optimized AI workload execution.

Retirement-plan servicing inefficiencies, primarily caused by cumbersome document retrieval processes, can be significantly improved through an AI-based platform that enhances accuracy, reduces call times, and ensures compliance. Traditional servicing methods involve customer service agents manually searching through multiple documents, leading to wasted time and potential audit risks. The proposed solution, known as retrieval-augmented generation (RAG), uses AI to create a streamlined, auditable workflow by breaking down documents into searchable sections, allowing for fast and precise retrieval of necessary information. This AI-enhanced process not only shortens call times and reduces costs but also ensures each response is backed by a complete audit trail, leading to stronger fiduciary control in retirement-plan servicing.

Vertesia addresses the infrastructure bottleneck faced by enterprises deploying large language models (LLMs) at scale by introducing a unique air traffic control pattern for AI agents. This approach allows agents to request clearance before making LLM calls, ensuring efficient use of dynamic quotas offered by providers like Bedrock and Vertex AI, which vary based on factors such as regional demand and system load. By employing a durable workflow architecture with Temporal and intelligent rate limiting, Vertesia enables agents to pause and resume operations without wasting resources, adapting to real-time capacity availability. This results in significant improvements in throughput and error rates during large-scale AI deployments, allowing for more predictable and efficient use of infrastructure resources. The platform's ability to dynamically discover and optimize capacity utilization reduces operational overhead and cost, enabling faster and more reliable AI processes without manual intervention.