Key Takeaways
PLM data is the dataset executives want AI agents to access most.
When Propel asked 400 senior leaders which system held the most valuable data for AI agents to act on, PLM topped the list, ahead of both CRM and ERP. This confirms that product data, not just customer or transactional data, is now central to enterprise AI strategy.
73% of leaders call MCP critical to their AI strategy, but adoption requires the right governance.
While 67% plan to implement Model Context Protocol within six months, roughly half of respondents cited security, data governance, or compliance as their top concern before moving forward. The gap between intent and execution is a governance gap, not a technology gap.
Propel MCP operates on live data with user-level permissions and a full audit trail, never a copy.
Propel MCP performs actions directly against the platform's live data using the requesting user's exact existing permissions, down to the field level, with every action automatically logged, rather than exporting data to an external store where context, access controls, and traceability are lost.
Teams and Slack are becoming the default surface for AI-driven product data queries.
In live polling during the session, 59% of attendees said they interact with AI through Microsoft Teams and 41% through Slack, confirming that users want AI answers to meet them inside the collaboration tools they already use daily, not force them into a separate interface.
Starting with Read access before Write access builds the trust needed to scale AI use cases.
Best practice is to start AI implementations with basic Read and Summarize actions before expanding to Write access, so organizations can validate real user demand before granting agents permission to change product data.
Saved skills turn one-time AI prompts into repeatable, deterministic workflows.
As shown in the demo, Propel MCP is able to convert a multi-step BOM cost reconciliation process between Propel PLM and an ERP system into a portable skill file, which locks in the exact steps and rules so future requests produce consistent results while using fewer tokens.
Propel’s 9.80 release increases token efficiency of AI queries by 30 to 50%.
Propel’s latest platform release improves how efficiently tokens are used when third-party AI assistants query Propel, and that this inbound capability comes at no added cost for licensed customers.
