Applying Agentic AI to workforce management
Use cases ideated
10+ use cases ideated
Time to Pilot
9 weeks to production pilot
Customer Engagement
5 customers in production pilot
Business Impact
Significant task time reduction
About our client
Our client is an industry-leading workforce management platform that simplifies scheduling, time and attendance, communication, and compliance for businesses worldwide, empowering managers and hourly employees across various sectors.
The challenge
Our client aimed to leverage the power of Agentic AI to unlock significant operational efficiencies and introduce new, intelligent features into its platform. The primary challenge was the broad scope of potential use cases, ranging from hiring to payroll, which required a focused effort to quickly identify and validate a single high-impact pilot feature. The goal was to identify a use case that delivered immediate customer value while demonstrating the technical feasibility of a scalable AI architecture.
The solution
Mantel were chosen as a trusted partner to tackle this challenge. We executed a rapid, focused, nine-week engagement:
- Discovery and prioritisation
We facilitated a targeted workshop to evaluate more than 10 potential Agentic AI use cases. We ultimately selected an Intelligent Timesheets and Scheduling Agent as our optimal pilot. This moved beyond simple summarisation to complex action execution, enabling the AI to handle high-intent requests such as “Find a replacement for this shift,” “Approve all of yesterday’s timesheets,” or “Create a roster for the kitchen tomorrow.” - Architecture and development
A cross-functional team designed and built a working Proof of Concept (POC) on a scalable AWS cloud foundation. The core of this solution was a multi-agent system integrated with an MCP (Model Context Protocol) Server. - Strategic design
The approach centered on creating a robust, API-driven architecture and designing a foundational Natural Language Interface (NLI). Crucially, the solution was decoupled from Deputy’s core codebase, ensuring maximum flexibility for scaling and future innovation.
The outcome
The engagement successfully met the goals for rapid innovation and strategic platform development:
- Validation: The pilot successfully validated the agentic workflow in a live environment, proving the system could accurately interpret and execute complex scheduling and timesheet commands.
- Market launch: The new AI capability is currently in closed beta testing, and is scheduled to go public in Q1 2026.
- Platform delivery: Beyond the solution itself, the engagement delivered 3 strategic assets:
- Validated use cases: A prioritised backlog of high-impact opportunities.
- Foundational AI platform: A reusable “AI engine”, ready for future expansion.
- Strategic roadmap: A clear path for subsequent AI initiatives.
- Future-proofing: This rapid innovation established a reusable multi-agent architecture that will accelerate the development of future Agentic AI features and applications across our client’s entire ecosystem.