athenahealth Deploys Agentic AI to Patients
Healthcare technology company athenahealth has launched AI-powered, agentic patient communication tools across its provider network, which serves one in five Americans. The new tools automate patient engagement tasks like reminders and scheduling. The rollout provides a model for integrating analytics-driven insights directly into operational healthcare workflows.
- The new "Patient Conversations" feature includes two-way texting and a virtual front desk assistant with multilingual support for inbound calls, with future plans for real-time language translation. - An AI-powered "Waitlist Scheduling" feature will automatically identify openings from cancellations and offer them to patients via text message to optimize schedules. - The system is built on athenahealth's "Advanced Intelligence Layer," which synthesizes data from multiple sources like other EHRs, payers, and registries to provide actionable insights directly within clinical workflows. - This agentic AI is designed to be more autonomous than traditional rules-based automation, enabling it to interpret context, make decisions, and perform multi-step tasks like rescheduling and notifying care teams without manual intervention. - Beyond patient communication, athenahealth is also developing AI tools like "athenaAmbient," a digital scribe that automates clinical notes, and "Sage," an AI copilot for querying patient charts in real-time. - The company is piloting a Model Context Protocol (MCP) server, which allows external AI models to communicate with the athenaOne platform, enabling partners to build custom AI agents using diverse data sources. - The athenahealth Marketplace offers integrated, specialized agentic AI solutions from partners like Aidion Health for voice AI and mdhub for automating mental health patient admissions. - This initiative is part of a broader strategy to create an "AI-native" athenaOne platform, aiming to reduce administrative tasks, which clinicians report spending an average of 15 hours per week on.
