451 IT Insider: A roundup for IT decision-makers (July)

Acting as a universal adapter, MCP uses a host-client-server architecture, enabling LLM-based software to receive information and instructions from multiple databases, APIs or services. MCP clients can be created by an MCP host (typically an AI application or development environment connected to an LLM); lightweight MCP servers maintain 1-1 connections between clients and information sources, prompts and actions, exposing specific capabilities and giving context to the host application and the LLM. MCP servers are developed and maintained by the companies and communities that build them.

Businesses can now build effective AI agent frameworks so systems can operate autonomously to accomplish complex tasks. The most successful agentic systems are built with simple, composable patterns.

"The return on investment for it [GenAI] is phenomenal. Being able to save hours of work or days or weeks or months of labor to have something, even if it's just a base template designed for you within seconds, is perfect. It is extremely helpful."

-IT/engineering manager/staff, undisclosed employee count and revenue, other industry.
Source: 451 Research's Voice of the Enterprise in-depth interview, February 2025.

Key industry conferences: Quantum.Tech USA and Shoptalk

Quantum computing acceleration

This year's Quantum.Tech USA reflected renewed momentum behind quantum technology. Beyond highlighting quantum's progress to date, the conference focused on remaining challenges for the still-young industry. The urgency for continued research and investment in quantum technologies was underscored by a focus on national quantum initiatives and a full day dedicated to quantum's impact on cybersecurity. At the same time, many keynote sessions also discussed the need for a more robust talent pipeline to support the industry's anticipated growth. Our Quantum Computing Market Monitor & Forecast published in June 2025 reveals growth in spending as pilot projects come to fruition. An inflection point is targeted for 2030 with the emergence of fault-tolerant quantum systems.

In the near term, generative AI approaches will likely be helpful in the continued development of quantum computers, such as by accelerating the progress of quantum hardware or improving the software interfaces between quantum machines and classical systems.

Figure 2: Quantum computing business impact vs. intent to invest

Intent to invest in quantum computing lags anticipated business impact, and even those who anticipate quantum computing will have a significant business impact in the near term are not always planning to invest in the technology immediately. This reflects the market's immaturity and should be a call to action to suppliers that more education is required. CTOs should stay on top of the emerging technology for improvements in enhanced problem-solving that can create a strategic advantage by allowing companies to innovate faster, optimize operations and improve decision-making.

Balancing personalization and privacy

Companies are aware of the massive impact that AI will have on areas such as personalization and operational assistance, a topic of discussion at the Shoptalk conference. Commerce's more involved technology players are also identifying additional areas of interest, such as AI-driven improvements to spur product discovery on search engines and relevant sites. Our Voice of the Enterprise: Customer Experience & Commerce, Merchant Study 2025 provides insight into how merchants' IT, customer and digital experience teams must navigate a complex landscape transformed by both agentic and generative AI as well as unified commerce. These technologies promise to reshape customer engagement and operational efficiency, demanding strategic agility and robust data strategies that may be costly but will reward brand practitioners in the long run.

Retailers that effectively integrate these advancements will be well-positioned to deliver personalized experiences and maintain a competitive edge in an increasingly complex digital ecosystem. Consumers are shopping online and in person, seeking more seamless connections between their experiences. To strengthen their relationships with consumers, brands must increasingly meet customers where they spend their time, which requires the right blend of data and media. Increasing consumer acceptance of personalization, balanced with privacy, will fuel momentum for first-party data.

Longitudinal data from our Voice of the Connected User Landscape: Connected Customer, Trust & Privacy surveys shows an evolutionary shift toward balancing personalization and privacy.

While privacy remains important, our 2024 survey marked a dramatic shift in which consumers' preferences for personalization versus privacy shifted more heavily toward personalization. Between our 2023 and 2024 surveys, we recorded a 25-percentage-point decrease in preference for privacy and a 16-point increase in preference for personalization. Emphasis on personalization is primarily driven by Gen Z and millennial preferences, suggesting a long-term trajectory toward balanced personalization-privacy expectations.

Figure 3: Shifting balance between preferences for privacy and personalization

Emerging tech: Edge infrastructure investment driven by AI and CX

Enterprises are deploying edge infrastructure and services to enhance customer experience, optimize operations and unlock previously inaccessible use cases. Enterprise sentiment points to a rapidly maturing edge computing landscape, where infrastructure is deployed across diverse environments and powered by both traditional and emerging technologies. AI is catalyzing investment, use cases are expanding into customer-facing and predictive domains, and spending is rising sharply — even as cost, security and performance remain persistent barriers.

CTOs must understand the use cases to allocate funds across the business as spending increases properly. A strong 68% of organizations report increased edge spending over the past year. Notably, 47% added to their edge budgets to meet growing AI requirements, while 38% reallocated funds from other projects to support edge AI initiatives. This momentum signals that edge is becoming a strategic budget priority, particularly as enterprises shift AI workloads from centralized cloud environments to the edge for faster, more context-aware processing.

"I think [our usage of edge compute] will slowly creep up only because … we're very much responsive to what our business needs are and how we're able to deliver them. There's an increased need to gather so much [data] … [But] the core of our processing occurs in the public cloud."

-Mid-level management, 10,000-49,999 employees, $10B+ revenue, consumer retail.
Source: 451 Research's Voice of the Enterprise in-depth interview, July 2024.

Customer experience is a top use case today. Edge computing increasingly drives business outcomes, with 39% of organizations citing customer experience optimization as a use case. Other leading applications include private 4G/5G network enablement (36%) and asset tracking or monitoring (32%). This shows that edge is not just about IT efficiency — it's about enabling real-time, data-driven services that directly impact customer satisfaction and operational agility. Looking ahead, 24% of organizations expect to support marketing/advertising automation and sales/demand forecasting through edge deployments.

M&A in focus: Qualcomm hands over $2.2 billion for AI silicon intellectual property specialist Alphawave

Qualcomm is pitching its Oryon CPUs and Hexagon neural processing units as AI inferencing engines. It views Alphawave's capabilities as key assets that will improve its portfolio's performance and power efficiency as demand for AI inferencing is exploding and datacenters are transitioning to custom CPUs. Advanced processors are critical for enabling the performance and efficiency of modern technology infrastructure, particularly in areas such as datacenters, AI and 5G. AI-specific silicon technology, such as that developed by Alphawave, allows data to travel faster, more reliably and with lower power consumption, addressing exponential data growth and the increasing demands of compute-intensive applications. This technology can help CTOs meet the rising demand for computing performance and efficiency.

10 companies to track

Defined.ai specializes in AI training data and promotes ethical practices in AI development. By maintaining a commitment to transparency, data privacy and adherence to regulatory standards, Defined.ai supports organizations in creating AI systems that are effective and aligned with responsible data practices, providing ethically sourced and curated datasets.

Hex is expanding from data science to DIY analytics for business users. It provides a composable workspace for data teams to conduct data science by coding predictive models in Python, and it also enables users to write SQL queries. With interchangeable Python and SQL support, data scientists and analysts can choose the most appropriate method for a specific use case.

Meibel aims to simplify the development and management of generative AI capabilities and help organizations detect and resolve low-quality outputs. As part of a platform encompassing AI capability development, data ingestion, evaluation and model routing, it focuses on auditability and evaluation, equipping teams with tools to pinpoint where issues arise, assign confidence scores to outputs and take corrective action.

Sonar anticipates a coming time when AI will generate more code than humans, and enterprise strategies for building and managing software will need to adapt. Far from replacing human developers, machine-generated code requires proactive supervision to ensure that it is high-quality, maintainable and secure in a business context. Sonar is taking a developer-first approach to the challenge, integrating static code analysis, policy enforcement and issue remediation at the start of the software life cycle.

Swimm accelerates legacy and mainframe modernization projects by helping software organizations understand their own code. It is building intellectual property specific to legacy programming languages such as COBOL and PL/I to transform those languages' business rules and logic flows into AI-ready primitives for modern deployment. After extracting language-specific concepts and transforming them into digestible, editable snippets, the platform uses an LLM to explain what they do in natural language, creating a "knowledge layer" of code understanding and exposing it to different applications and agents.

Orum has quietly achieved something few fintechs have: a direct connection to the Federal Reserve's payment rails, a level of access typically reserved for banks. This also gives Orum a pricing edge based on operational efficiencies and reduced reliance on bank partners and third-party processors. The company's always-on capability is especially valuable for platforms serving users across time zones or needing to push funds on weekends. By abstracting the messiness of legacy banking rails and wrapping the capability into a developer-friendly API, Orum positions itself as a logistics layer for modern money movement.

SkySQL is doubling down on its agentic AI strategy, which consists of built-in and user-created agents that help developers and database administrators extract maximum value from their data stored in the SkySQL database. SkyAI Agents can be predefined, user-created or semi-autonomous. These agents range from helping database administrators manage their SkySQL database environments to assisting developers to infer meaning from their data schema, and all agents allow users to interact using natural language.

MagicOrange's product offering delivers a unified cost and profitability platform, which brings transparency and line-of-sight visibility into profit, cost and resource allocation. Its reporting, analytics and insights are tailored to help specific roles and personas make data-driven decisions. Its key use cases are data sprawl, eliminating waste, profit optimization, producing a clean bill of IT health, and guiding digital transformation. Its key features are cost modeling, cloud cost management, budgeting and forecasting, reporting and predictive modeling.

CloudHedge is an AI-based automated refactoring and containerization technology for enterprise cloud migration and modernization. The startup's key proposition is datacenter virtual machine-to-cloud container migration, principally shifting Linux-, Windows- and AIX-based application servers and databases into cloud containers. Its container-first approach offers Docker file and YAML/Helm chart auto-generation. License costs are covered under Amazon Web Services' independent software vendor funding program for VMware Exit and Windows Workloads.

Vectara emerged in 2020 as a serverless AI-powered semantic search platform developer. As generative AI gained momentum, the company carved out a niche in the emerging "trusted AI" field, leveraging its expertise in model development and neural search techniques to target retrieval-based workflows. As enterprises move from experimentation to production, they become increasingly aware of the architectural intricacies behind generative AI applications, particularly those leveraging internal data.

What to watch

Companies must strategically navigate these technological trends to maintain a competitive edge and drive innovation. By staying informed and investing in emerging technologies, businesses can capitalize on the opportunities presented by quantum computing, AI and edge computing, while also addressing evolving consumer expectations regarding personalization and privacy. These developments will play a crucial role in shaping the future of AI and enterprise technology, offering new opportunities and challenges for businesses worldwide. S&P Global Market Intelligence 451 Research provides essential insight into the pace and extent of digital transformation across the global technology landscape, backed by a global team of industry experts. Keeping an eye on these developments will be essential for organizations aiming to navigate the complexities of digital transformation successfully.