Assembled launched a Model Context Protocol (MCP) server for contact center workforce management (WFM) that connects tools like Claude and ChatGPT to live operations data.
The company said the launch makes it the first WFM platform in the contact center category to ship an MCP server. Assembled positioned the move as a way for workforce teams to diagnose performance, run scenarios, and take intraday actions through natural language prompts rather than dashboards and custom reporting.
Assembled also used the announcement to push back on the idea that AI will shrink the contact center workforce. The company cited a16z reporting that the Philippines contact center industry grew from 1.15 million workers in 2016 to 1.9 million in 2025. It also cited projections of another 70,000 jobs in 2026.
Why Assembled Thinks WFM Is Shifting to Agentic Front Ends
Assembled framed MCP as a practical bridge between AI assistants and workforce data that usually sits behind APIs and integrations. Ryan Wang, Co-Founder and CEO at Assembled framed the move as a practical shift:
“The future of WFM is agentic. Instead of waiting hours for an analyst to pull a report from one tool and stitch in data from three others, leaders will just ask their AI assistant to pull the answer from all of their data connections in one conversation. Instead of scenario planning taking a week, leaders will run any ‘what if’ in a single conversation.”
The company’s timing also lines up with broader enterprise UX forecasts. Gartner estimates that by 2028, a third of user experiences will shift from native applications to agentic front ends.
Assembled said customers are already using the MCP server for tasks that blend reporting and action. Examples in the release included breaking down SLA misses by channel, comparing forecasted versus actual volume by hour, and publishing short overtime windows for phone coverage.
Assembled also positioned its existing product approach as a fit for MCP-style workflows. It described itself as an open platform with a developer-friendly API. It also said power users already build custom agents, integrations, and automated workflows on top of its foundation.
The company argued that MCP expands that flexibility to non-technical users. It said leaders can now build cross-tool systems without code, custom integrations, or API keys.
‘Democratizing’ WFM Data Beyond The Ops Team
Assembled’s announcement leaned heavily on the idea that WFM can become more proactive when routine reporting moves faster and closer to decision makers.
In an assessment, Christian Shrader, Senior Manager, Operations Tools and Systems at Checkr Inc. highlighted:
“What excites me about the Assembled MCP isn’t just that I can move faster on analyses that used to take days. It’s that it democratizes WFM data across the business. Leaders no longer need to wait for a dashboard or an analyst. They can ask the questions where they’re already working, in a Claude window, and get an answer grounded in real operational context.”
Assembled said the MCP server uses pre-aggregated metrics that incorporate each customer’s business logic. It also said the product provides read and write access across forecasting, scheduling, intraday operations, workforce composition, compliance, and performance analytics.
On security and access controls, Assembled said all queries and actions are authenticated via OAuth and scoped strictly to individual accounts, with no cross-customer data access.
It’s easy to focus MCP talk on agent assist and front-line automation. But WFM is where contact centers decide how work gets staffed, prioritized, and governed. If MCP becomes common in workforce tooling, more of that decision making could move into conversational workflows that run faster than traditional reporting cycles.
In the near term, the big test is practical. WFM leaders will look for fewer hours spent stitching together exports and dashboards, faster scenario planning, and tighter intraday control. Over time, the bigger shift may be cultural. If the interface for workforce decisions becomes the same AI assistant window leaders already use, the expectation for real-time operational answers won’t stay optional for long.
Join the conversation: Join our LinkedIn community (40,000+ members): https://www.linkedin.com/groups/1951190/
Get the weekly rundown: Subscribe to our newsletter: http://cxtoday.com/sign-up
