Healthcare Innovation lately wrote about a presentation by Christopher “Topher” Sharp, M.D., chief medical info officer at Stanford Well being Care, about efforts there to convey synthetic intelligence into medical use safely, ethically and cost-effectively.
In one other instance of how innovation round AI is evolving, Wealthy Mendola, enterprise CIO for Johns Hopkins College and Johns Hopkins Drugs, lately launched a problem for IT workforce members to assist Johns Hopkins undertake and leverage AI. Summaries of some of the submissions made to date by 16 workforce members and revealed on the IT@JH web site vary from instruments for coding, finding out, and medical documentation to options that might scale back workload and enhance enterprise assist. They provide a glimpse of the kind of AI innovation effervescent up at many educational medical facilities and neighborhood well being methods throughout the nation.
Very similar to Stanford Well being Care created a safe setting known as Safe GPT the place management might permit for full experimentation by the entire group, HopGPT (previously Hopkins AI Lab) gives members of the Johns Hopkins neighborhood safe and easy accessibility to Massive Language Fashions (LLMs) from firms similar to OpenAI, Anthropic, and Meta. Developed by IT@JH, this safe setting mitigates safety and privateness dangers whereas making certain that knowledge stays protected.
Within the AI/HopGPT Problem, staff are submitting a abstract of how they utilized HopGPT or one other AI instrument so as to add worth or effectivity to a physique of labor inside IT. This problem offers a possibility for particular person contributors to achieve visibility, be artistic, and have an effect each inside and out of doors the scope of their common roles.
Though the AI/HopGPT problem deadline has been prolonged to March 2, listed below are the chief summaries of three highlighted proposals:
Sightline FAQ Agent, Alex Quigley & Olga Sankova, Sightline
Sightline is a enterprise transformation program bringing Workday to Johns Hopkins College and Johns Hopkins Well being System. Workday will even assist Johns Hopkins with a unified platform for self-service and administrative duties in 2027. The workforce designed and examined a Microsoft Copilot agent utilizing historic program responses and vetted information sources. The agent saves time spent manually digging by spreadsheets or repetitively asking purposeful specialists for solutions. It additionally creates one constant voice coming from many Sightline representatives to strengthen and professionalize responses. The Sightline FAQ agent is a template for dealing with Q&A with different complicated initiatives supported by massive groups.
Medical Documentation Enchancment Scalable AI Undertaking, Ebony Williams, Well being IT
Incomplete and fragmented medical documentation in Epic contributes to affected person security danger, coder and biller inefficiency, misplaced income, and excessive audit findings. This proposal encompasses a phased, AI-enabled documentation assist layer embedded immediately in Epic. This resolution is designed to enhance documentation completeness and readability in real-time, scale back clinician burden, and make sure the narrative helps correct coding, billing, and regulatory necessities with out making autonomous choices or disrupting care. This strategic funding safeguards income, improves compliance, and creates a scalable platform for ongoing innovation.
Software program Asset Administration & Governance Engine (SAGE), Nitin Ale, Enterprise Enterprise Options
There isn’t a single system of report for software program property throughout IT or the enterprise. SAGE (Software program Asset Administration & Governance Engine) is an AI-assisted intelligence layer that consolidates fragmented software program knowledge right into a single determination view. It doesn’t alter governance or automate approvals — its function is to supply clear, constant context so groups could make better-informed choices about buying new software program options.
After March 2, the IT management workforce will consider all of the purposes they’ve obtained and decide subsequent steps for essentially the most extremely rated concepts.
