Research
We conducted in-depth research with NHS Trust administrators to understand staffing workflows, communication gaps, and compliance requirements before architecting the Agentic AI framework.
We developed an Agentic AI system powered by AutoGen and Claude Sonnet 4.5 to automate shift fulfillment for NHS Trusts. The solution intelligently matches unfilled shifts with suitable staff based on skills, preferences, and availability, and autonomously engages them through chat to confirm assignments — ensuring faster, smarter, and goal-driven workforce allocation.
NHS Trusts across the UK face persistent challenges in filling last-minute or unassigned shifts due to complex scheduling, varying staff availability, and communication bottlenecks. Manual follow-ups and delayed responses from medical staff often result in operational inefficiencies, increased costs, and compromised patient care coverage. The goal was to automate and intelligently manage this end-to-end process — from identifying open shifts to engaging suitable staff — while maintaining compliance and a human-centric approach.
Agentic AI ecosystem leveraging AutoGen and Claude Sonnet 4.5 LLM — a dynamic, self-coordinating system of AI agents designed to autonomously manage NHS shift allocations.
Vacancy Management Agent — retrieves all open shifts for the upcoming week, analyzes job requirements, and matches them with the most suitable staff based on skills, certifications, availability, and preferences.
Engagement Agent — automatically initiates personalized chat sessions with staff, suggesting relevant shifts and handling their responses in real time.
Supervisor Agent — oversees the process, ensuring alignment with organizational goals, fairness in staff allocation, and timely closure of vacancies.
The system continually refines its recommendations using staff feedback, response time patterns, and previous shift performance data.
Integrated seamlessly with NHS scheduling and communication platforms to ensure smooth deployment and minimal disruption.
We conducted in-depth research with NHS Trust administrators to understand staffing workflows, communication gaps, and compliance requirements before architecting the Agentic AI framework.
Developed and trained three autonomous AI agents using AutoGen, integrated with Claude Sonnet 4.5 for natural conversation and context retention, ensuring real-time responsiveness and accuracy.
The solution was deployed within NHS’s existing scheduling infrastructure, running securely in a cloud-native environment with continuous monitoring and performance optimization.
The Agentic AI system transformed the way NHS Trusts manage shift allocations — automating over 80% of manual communication, reducing vacancy closure time by 60%, and ensuring fairer staff distribution based on skill alignment and preferences. This initiative not only improved workforce satisfaction but also enhanced operational continuity and patient care quality — setting a new benchmark for intelligent workforce automation in healthcare.
