For decades, manufacturers have lived with a frustrating paradox. The enterprise software vendors they rely on — the Siemens, PTCs, and Dassault Systèmes of the world — promised digital transformation. What they often delivered instead was a walled garden: a tightly integrated ecosystem that worked beautifully within their own product family and created friction at every boundary with the rest of the enterprise.
The result is a system landscape that most manufacturers know intimately: a PLM system that doesn't talk to ERP without a costly integration project, a quality management system siloed from engineering, CAD tools requiring manual handoffs at every software boundary, and a CRM that rarely has an idea of what version of the product the customer actually purchased. The data exists. It just doesn't flow.
That dynamic is changing — faster than most of the industry has processed. AI agents are entering the enterprise. And the infrastructure being built to support them is rendering the walled garden model not just outdated, but obsolete.
The Protocol That Changes Everything
In November 2024, Anthropic released the Model Context Protocol (MCP) — an open standard that defines how AI systems connect to external tools, data sources, and business logic. Think of it as the USB-C port for AI: a universal interface that eliminates the need for one-off integrations between every AI model and every enterprise system.
Before MCP, connecting an AI agent to your enterprise meant building custom connectors for every system, with unique authentication flows, data models, and error handling for each. The complexity scaled exponentially. MCP collapses it, letting any MCP-capable AI client connect to any MCP-capable server — once — instead of rebuilding the connection for every possible pairing.
The adoption has been remarkable. What React achieved in three years, MCP matched in sixteen months — hitting 97 million monthly SDK downloads by early 2026. OpenAI, Google, Microsoft, and Amazon all standardized on it. In December 2025, Anthropic donated the protocol to the Linux Foundation, cementing it as a neutral industry standard rather than any single vendor's property.
But raw adoption numbers don't capture what MCP actually means for manufacturers. The deeper implication is this: AI agents can now reach across every system in your enterprise — PLM, ERP, MES, CRM, supplier portals — and reason across that data in real time, without a multi-year integration program as the prerequisite.
The question is no longer whether AI will transform product operations. It's which platforms are built to make that transformation real, fast, and trustworthy.
Why Architecture Determines Who Wins
Not every PLM vendor is equally positioned to take advantage of this shift. Legacy platforms face a fundamental problem: their data is locked behind proprietary models, often on-premise or in first-generation cloud architectures that weren't designed for the open, composable world MCP enables. Building MCP support on top of those systems is a multi-year engineering effort — and the trust infrastructure required to do it safely at enterprise scale is not something you can bolt on quickly.
Propel was built differently. As the only platform combining PLM, QMS, and PIM in a single unified data model, Propel's product records, quality records, BOMs, change orders, and compliance documentation are all connected — not integrated across seams, but genuinely unified. When MCP-based connectivity and agentic AI capabilities become available, Propel inherits them immediately across all of that data. There is no translation layer. There is no integration backlog. The product thread is already structured, governed, and ready for AI to act on.
This architectural advantage is not theoretical. Propel closed its best fiscal year in company history in early 2026, reporting 42% year-over-year growth in bookings, driven by manufacturers migrating from legacy on-premise and first-generation cloud PLM systems. Customers are already seeing up to 40% reduction in change order cycle times, 100% data accuracy, and measurable improvement in supplier collaboration — results that become the foundation for the next layer of AI capability, not a distant aspiration.
Propel One: Agentic AI That's Already Working
Propel One is Propel's agentic AI suite — and it is operating on live product and quality data today, not in a demo environment or a pilot program.
Built on enterprise-grade AI infrastructure with role-based permissions and full audit trails governing every action, Propel One is deeply embedded intelligence running on structured, trusted data across PLM and QMS workflows. It is not a chatbot layered over a legacy system. It is an AI layer that understands the product record — items, BOMs, change orders, quality records, supplier data, training documentation — and acts on it with precision.
In change management, Propel One pre-populates reviews with impact summaries, affected component owners, historical precedents, and customer contract references. What used to require days of manual context-gathering becomes a focused 15-minute decision.
In quality workflows, agents proactively surface CAPA risk patterns, flag component obsolescence from real-time supply chain data, and automate compliance documentation. For companies operating under FDA QMSR, ISO 13485, or EU MDR requirements, that kind of proactive, auditable quality intelligence isn't a nice-to-have — it's a competitive necessity.
Early Propel One customers — including Allegro MicroSystems, Guardant Health, and Breg — are already running agentic workflows that would have required multi-year integration projects on legacy platforms. Guardant's VP of Quality Systems described the impact simply: their teams can now make smarter decisions, faster, at every level of the organization.
As MCP adoption deepens across the enterprise software ecosystem, Propel One agents gain the ability to reach beyond Propel itself — pulling ERP inventory and cost data, cross-referencing MES as-built records, surfacing supplier quality data — all within a governed, permission-aware framework. Integration investments manufacturers have already made between Propel and systems like SAP, NetSuite, or Oracle become AI-accessible without rebuilding them. The existing integrations become an AI asset.
The Window Is Open — But Not Forever
The PLM market is in the middle of a genuine platform shift. The vendors that get ahead of it aren't the ones with the longest feature lists or the biggest installed bases — they are the ones whose architecture was built for the world AI is creating, not the world that preceded it.
For manufacturers still asking when AI will be practical for product operations, the answer has arrived. The walled garden was never the right architecture for how manufacturing enterprises actually work. The open, connected, agent-ready enterprise is the architecture that wins. Propel is executing on it now.
.png)
