Top 10 AI Prompts and Use Cases and in the Healthcare Industry in Puerto Rico

Puerto Rico's healthcare scene is in a fast‑moving pivot: local surveys show 84% of organizations are already using AI in at least one function, yet 59% cite a lack of in‑house expertise as the main adoption barrier - a skills gap made starker by chronic funding shortfalls and workforce loss (the island still averages about one doctor leaving per day after Hurricane Maria disrupted care and records).

Practical wins - from cloud interoperability projects that stitch patient histories back together to AI tools that ease clinician paperwork and boost patient follow‑up - are already underway.

For leaders planning pilots, the V2A State of AI in Puerto Rico report offers the adoption snapshot, AWS's account of rebuilding Puerto Rico's health infrastructure traces why interoperability matters, and Nucamp AI Essentials for Work bootcamp - 15‑week practical AI training provides a hands‑on path to close the skills gap quickly.

Metric	Value
Local AI adoption	V2A report: 84% of local organizations using AI
Top barrier	59% lack in‑house expertise
Nucamp AI Essentials	Nucamp AI Essentials for Work - 15 weeks (early bird $3,582)

“We have been exploring and experimenting with AI for several years now – monitoring carefully and responsibly how AI can effectively support the patient care experience.” - Marc D. Miller, President and CEO, Universal Health Services

Selection balanced practical payoff for Puerto Rico with patient‑safety and adoption risk: choices were informed by a recent scoping review of AI threats to patient rights and safety - Archives of Public Health (2010–2023) and benchmarked against global deployment trends in NVIDIA's State of AI in Healthcare 2025 report to capture which use cases are actually moving from pilot to production.

Priority criteria included documented clinical impact, measurable operational ROI, privacy and safety risk, ease of local workforce adoption, and interoperability with existing systems - practical filters especially important given Puerto Rico's staffing and funding constraints.

Real‑world evidence mattered: a local retrofit highlighted in our research showed a 70% jump in detection accuracy while slashing false positives, underscoring why proven pilots informed the list (Amgen Juncos AI retrofit case study: detection accuracy improved by 70%).

The result is a Top 10 that favors safe, scalable, and workforce‑friendly AI first - pilots that can be stood up quickly and evaluated against clear safety and efficiency metrics (practical AI deployment guidance for Puerto Rico healthcare systems).

For Puerto Rico's hospitals and neurology clinics facing tight staffing and rising demand for MRIs, icometrix's FDA‑cleared icobrain suite offers a practical way to turn complex brain MRI series into clear, quantitative reports that speed decision‑making: icobrain integrates with PACS/EMR via an on‑site icobridge for secure DICOM routing and can deliver results in minutes, while peer‑reviewed evidence shows as much as 150% increased detection of MS disease activity, a 40% boost in reading speed and 2.5× faster detection of treatment failure - metrics that matter when follow‑up windows and specialist time are limited.

The platform also supports disease‑specific workflows (MS, Alzheimer's, ARIA monitoring for amyloid therapies, TBI, epilepsy) and now sits on a clearer path to billing after the AMA issued CPT® III codes for AI quantitative brain MRI analysis, which helps create reimbursement pathways for Puerto Rico providers considering pilots.

With GE HealthCare's announced intent to acquire icometrix, local leaders should watch for tighter MRI‑to‑report integrations and scaled access to icobrain aria for monitoring anti‑amyloid therapy safety as they plan pilot projects that balance clinical impact, workflow fit, and reimbursement prospects (see icometrix for product and integration details).

Metric	Value / Source
Increased MS activity detection	150% (icometrix clinical evidence)
Radiology reading speed	+40% (Van Hecke et al., 2021)
Faster detection of treatment failure	2.5× faster (Sima et al., 2021)
Turnaround time	Minutes (icometrix clinical workflow)
Reimbursement pathway	AMA CPT® III code for AI brain MRI quantification

“Precision and personalized medicine are indeed becoming the standard of care quickly in neurology.” - Dr. Joseph Fritz

K Health's AI‑powered chatbot offers Puerto Rico a practical, high‑availability front door for primary, urgent and mental‑health needs - an always‑on symptom triage and history‑collector that can act like a digital triage nurse in the palm of a patient's hand at 2 a.m., freeing scarce clinic time for higher‑acuity care.

Built from millions of anonymized visits and already used by about 3.1 million patients, the system analyzes symptoms, suggests likely conditions and flags when human escalation is required; a retrospective study found clinicians agreed with K Health's results in over 85% of visits, and partnerships with Cedars‑Sinai and Mayo Clinic show the platform's ambition to embed AI into clinical pathways and personalized treatment algorithms.

For Puerto Rico leaders weighing pilot options, K Health's model illustrates how chatbot triage can cut administrative burden, extend access in rural areas, and support task‑shifting - see more on K Health's approach and the broader case for medical chatbots and ROI in healthcare.

Metric	Value / Source
Users	~3.1 million patients (K Health)
Clinician agreement	>85% match in retrospective study (K Health)
Primary use cases	Symptom triage, primary/urgent care, mental health, pediatrics (K Health)

Puerto Rico's clinics and telehealth programs can get immediate traction from Teladoc Health's push to fold Microsoft Azure OpenAI, Azure Cognitive Services and Nuance's Dragon Ambient eXperience (DAX) into the Solo telehealth platform - ambient note capture and automatic visit summarization reduce the grief of documentation that drives clinicians away (physicians spend roughly two hours on paperwork for every hour of care), free bedside staff for hands‑on work, and enable hybrid workflows like virtual nursing that are especially useful on the island.

For Puerto Rico's bilingual population, Microsoft's clinical tools also support multilingual encounter capture - conduct a visit in Spanish and generate English clinical notes - making continuity across specialists and payers smoother.

Leaders planning pilots should weigh easy integrations with existing virtual platforms and the promise of faster, more consistent after‑visit summaries and handoffs to community clinicians as practical levers to expand access and cut burnout: see the Teladoc Health announcement on Azure OpenAI and DAX integration and Microsoft's Dragon Copilot and Nuance Dragon Ambient eXperience overview for technical and workflow details.

Feature	Puerto Rico relevance / benefit
Ambient documentation (Nuance DAX)	Reduces clinician admin time; lets clinicians focus on patients
DAX Express + GPT‑4	Faster, accurate reporting to virtual and community clinicians
Multilingual encounter capture	Supports Spanish visits with English documentation for referrals and billing
Virtual nursing & hybrid workflows	Frees bedside nurses for hands‑on care, improving staff satisfaction

“Administrative burden and staff shortages are major reasons why clinicians are leaving the profession. We are focused on using AI to reassert and build the doctor‑patient relationship at a time when technology frequently does the opposite.” - Dr. Vidya Raman‑Tangella, Chief Medical Officer, Teladoc Health

Med‑PaLM 2 is a medicine‑tuned large language model that has been in limited testing at the Mayo Clinic since April 2023 and is now being previewed with select Google Cloud customers as a possible clinical assistant for tasks that matter in Puerto Rico - summarizing records, surfacing insights from unstructured notes, and trimming administrative load like prior authorization - while also answering complex medical questions with strong exam‑style performance (it was the first LLM to reach “expert” performance on MedQA and to pass MedMCQA).

Those features could be a practical boon for island clinics coping with physician shortages and heavy paperwork, but local pilots should pair potential efficiency gains with strict privacy and safety guardrails: Google emphasizes encrypted data controls in its Med‑PaLM 2 preview and Mayo Clinic–led testing aims to study real‑world limits and bias.

For Puerto Rico health leaders, the takeaway is clear - Med‑PaLM 2 shows promise as a clinician‑augmenting tool, not a replacement, and early, measured pilots tied to evaluation metrics will be essential (see Google Cloud's Med‑PaLM 2 overview and coverage of Mayo Clinic testing for details).

“Performance in a clinicopathological conference ‘in no way reflects a broader measure of competence in a physician's duties.'”

Northwell Health and the Feinstein Institutes offer a ready-made model Puerto Rico can adapt: combine remote physiological monitoring (RPM) - using blood pressure cuffs, pulse oximeters, scales and glucometers that stream data to care teams - with tele‑ICU, telestroke and consumer telehealth to keep more patients safely at home and catch deterioration earlier.

Their programs show how connected devices plus virtual specialists can shrink dangerous delays (about 2 million brain cells die each minute during a stroke) by routing bedside teams to remote stroke neurologists via telestroke, while eICU monitoring has been linked to roughly a 25% drop in ICU mortality in Northwell's early rollouts.

Northwell's PCORI‑funded home pulmonary rehab and diabetes/heart‑failure telemonitoring pilots - which paired exercise bikes and daily vitals uploads with remote therapists and educators - illustrate practical chronic‑care paths that could reduce avoidable hospital transfers across Puerto Rico's rural and island geographies.

For systems planning pilots, Northwell's telehealth program details and their enterprise collaboration with Teladoc Health show the integration points and scale options to evaluate next (see Northwell Health telehealth programs and the Teladoc Health collaboration for implementation notes).

Metric	Value / Source
ICU beds monitored (eICU)	180 beds across 13 ICUs (Northwell)
ICU mortality change	~25% decrease (preliminary Northwell data)
Telehealth surge	5,031% expansion Feb 2020–Feb 2021 (Northwell)
RPM devices used	Blood pressure cuffs, pulse oximeters, scales, glucometers (Northwell)

“Telehealth allows us to be proactive - reacting before there is a problem. It's a perfect blend of professional experience and technology, amplifying care for patients.” - Iris Berman, vice president of Telehealth at Northwell Health

Recursion's work with NVIDIA - anchored by a $50 million collaboration to scale foundation models and feed them through NVIDIA's BioNeMo platform - points to a new way Puerto Rico health leaders and research groups can accelerate preclinical discovery without buying a supercomputer: models and NIM microservices run in the cloud can let island teams test hypotheses, prioritize targets, and shrink costly wet‑lab runs.

The program leans on Recursion's vast proprietary dataset (over 23 petabytes and trillions of gene–compound relationships) and BioHive‑2's extreme compute (hundreds of H100 GPUs) to train models like Phenom‑Beta and community tools such as Boltz‑2 that promise much faster structure‑and‑affinity predictions; Puerto Rico institutions could tap these capabilities via cloud APIs rather than building local GPU farms.

For clinicians and system planners focused on rare diseases, oncology or accelerating local translational research, the practical “so what?” is clear: AI‑guided screening can reduce the number of costly experiments while surfacing better leads faster - read Recursion's announcement and explore NVIDIA's BioNeMo framework for how these services are packaged and deployed.

Metric	Value / Source
NVIDIA investment	$50 million (Recursion press release)
Proprietary dataset	~23 petabytes; trillions of gene–compound relationships (Recursion)
BioHive‑2 compute	Hundreds of NVIDIA H100 GPUs; TOP500 rank #35 (NVIDIA / Recursion)
Cloud platform	NVIDIA BioNeMo NIMs & APIs for generative AI in drug discovery (NVIDIA Docs)

“With AI in the loop today, we can get 80% of the value with 40% of the wet lab work, and that ratio will improve going forward.” - Ben Mabey, CTO, Recursion

For Puerto Rico's health systems grappling with limited data‑sharing capacity and strict privacy concerns, MLCommons' MedPerf offers a practical bridge: its federated evaluation lets algorithms be shipped to hospital servers for testing so patient data never leaves the premises, helping local clinical teams verify model fairness and generalizability against island populations without risky data transfers; MedPerf's orchestration can cut evaluation cycles “from months to hours,” speeding decisions about whether a tool is clinic‑ready, and recent work on a policy‑enhanced, smart‑contract workflow adds enforceable audit trails that give dataset owners more control and traceability.

That combination - neutral governance, faster federation, and enforceable policies - makes MedPerf a useful option for Puerto Rico pilots that need rigorous, privacy‑first validation before procurement or reimbursement discussions.

Learn more on the MedPerf open benchmarking platform, the MLCommons Medical Working Group project page, and the MedPerf smart-contracts March 2025 update for technical governance details.

Capability	Benefit / Source
Federated evaluation	Models run on local data; patient data never leaves the site (MedPerf)
Orchestration speed	Evaluates multiple models in hours vs. months (MedPerf / Nature paper)
Governance & privacy	Policy‑enhanced smart‑contract workflow (MLCommons, March 10, 2025)

“Transparency lies at the core of the MedPerf security and privacy design. Information security officers must know about every bit of information they are being asked to share and with whom it will be shared.” - Micah Sheller

Puerto Rico's finance and operations leaders can follow a clear, practical playbook: deploy targeted administrative automation to stop revenue leakage and free scarce staff for patient‑facing work.

AI‑driven coding, claims‑scrubbing, prior‑authorization automation and predictive denial analytics have already moved from hype to measurable impact - about 46% of hospitals now use AI in RCM and 74% report some form of revenue‑cycle automation, with documented wins like 50% fewer discharged‑not‑final‑billed cases and sharp coder productivity gains in real deployments (see the AHA market scan on AI in revenue cycle management).

Platform and vendor choices should prioritize secure, end‑to‑end workflows and measurable ROI - market research shows RCM leaders rank AI/GenAI investment as a top priority, driven by the need to stop denials that cost hospitals billions annually (learn more in Waystar's RCM trends brief).

For island systems, a phased engagement - start with eligibility, claim scrubbing and appeal automation, measure denials and cash‑flow improvements, then scale - keeps risk low and delivers visible savings that protect clinical capacity and patient access.

Metric	Value / Source
Hospitals using AI in RCM	46% (AHA market scan)
Hospitals implementing revenue-cycle automation	74% (AHA)
RCM leaders prioritizing AI/GenAI investment	92% (Waystar)
Estimated annual denial costs to hospitals	Billions annually (Notable / industry reports)
Prior-authorization denials reduced in one case	22% decrease (AHA case study)

“Cash is king.” - Matt Morgan, Vice President/CFO, Montage Health

Abarca Forward in San Juan has become a practical playbook for workforce augmentation, training and change management - mixing big‑stage strategy with hands‑on sessions (the popular “AI Petting Zoo” and co‑creation labs with Microsoft and Xtillion) so clinicians, pharmacists and operations teams can test tools before committing to pilots; that blend of experiential learning and executive alignment is exactly what Puerto Rico systems need to close the AI skills gap quickly and safely, and the event's track record - focused AI/PBM sessions in 2024, deeper AI‑PBM strategy in 2025, and plans to expand to Santurce in 2026 - keeps the conversation local and actionable.

Organizers report near‑perfect attendee satisfaction (100% of 2025 attendees would return), practical breakouts on specialty drug strategy and pharmacy as a consumer front line, and ongoing workshops aimed at turning pilot concepts into staffed, measurable programs - see the Abarca Forward program details and the 2024 recap for agendas and speaker highlights.

Item	Detail / Source
Hands‑on training	AI Petting Zoo & co‑creation sessions with Microsoft & Xtillion (Abarca Forward)
Attendee satisfaction	100% of 2025 attendees would return (Abarca Forward)
Local focus	San Juan events; 2026 moving to Santurce to deepen island engagement (Abarca Forward)

“AI is an existential opportunity and threat to every organization in healthcare, and now is the time to act.” - Bertil Chappuis, CEO, Xtillion

AI patient‑engagement tools offer a practical, measurable way to grow access and cut friction across Puerto Rico's clinics: vendors and case studies show AI chatbots and virtual assistants can cut missed appointments by up to 30%, boost medication adherence by about 30%, and drive dramatic workflow wins (Riseapps highlights the Black Doctor “Amina” assistant that accelerated diagnosis by 65% and Sensely's virtual assistant that cut 30‑day readmissions in some cases).

The market tidal wave is real - forecasts expect AI patient‑engagement to expand strongly over the coming decade - so the strategic question for island leaders is which pilots will win trust; public‑attitudes research finds reliability and transparency matter most to patients, more than cost or autonomy, so deployments should pair human oversight and clear audit trails with bilingual, literacy‑aware messaging.

That mix - targeted reminders, personalized education, conversational triage and analytics for marketing and service operations - can lower no‑shows, lift patient satisfaction, and give small systems outsized reach without huge capital spend (see Riseapps AI patient-engagement case studies, Polaris Market Research AI patient-engagement market outlook, and a JMIR survey on public attitudes toward AI in healthcare for evidence on trust drivers).

Metric	Value / Source
Appointment no‑show reduction	Up to 30% (Riseapps AI patient-engagement case studies)
Medication adherence uplift	~30% (Riseapps AI patient-engagement case studies)
Notable case example	Black Doctor Amina - 65% faster diagnosis (Riseapps AI patient-engagement case studies)
Market outlook	Forecast to USD 33.68B by 2032; CAGR ~20.9% (Polaris Market Research AI patient-engagement market outlook)
Public trust drivers	Reliability & transparency most influential on acceptance (JMIR survey on public attitudes toward AI in healthcare, 2025)

Puerto Rico health leaders should take a pilot-first path: pick one high-impact use case, run a short, measurable pilot with clear safety and equity checkpoints, then use a structured framework to move from experiment to scale - Perficient's “Start • Accelerate • Adopt • Scale” approach and its PACE and AI AMP toolkits offer a practical roadmap for that progression (Perficient PACE and AI AMP toolkits for healthcare AI pilots).

Governance and identity controls must be baked in from day one: consider identity-aware AI agents and access automation like SailPoint's Harbor Pilot to reduce risk while speeding workflows, and certify staff responsible for oversight with recognized credentials such as the IAPP Artificial Intelligence Governance Professional to keep policy, auditability and patient privacy front and center (IAPP Artificial Intelligence Governance Professional (AIGP) certification).

Finally, close the island's skills gap by pairing pilots with practical training - short, role-focused upskilling such as the 15-week Nucamp AI Essentials for Work 15-week bootcamp (registration) helps clinicians and operations teams write safe prompts, evaluate results, and turn pilots into sustained, measurable improvements in patient access and operational resilience.