Top 10 AI Prompts and Use Cases and in the Healthcare Industry in San Bernardino

San Bernardino County sits at a crossroads where rising demand, uneven access, and fast-moving tech converge: the county is home to 10 surgical hospitals within a state that counts about 153 such facilities, so smarter deployment of scarce resources matters (and fast) - which is why local agencies are leaning into AI for everything from predictive site analysis to automated workflows.

County leaders report concrete gains using enterprise tools (Microsoft Dynamics 365, GitHub Copilot) and GIS platforms to speed projects, cut coding time by roughly 30%, and turn maps into faster public-health action; read more about the county's digital strategy in the Insider GovTech article on San Bernardino's AI and GIS efforts.

Public-health teams already use Esri-powered dashboards to track outbreaks and target outreach, and generative AI/market-research tools can add visitation, traffic, and demographic layers to pinpoint where services are most needed - practitioners can build those practical skills in Nucamp's 15-week AI Essentials for Work bootcamp to safely apply prompts and tools across clinical and operational workflows.

Bootcamp	Length	Early-bird Cost	Learn More
AI Essentials for Work	15 Weeks	$3,582	AI Essentials for Work bootcamp - syllabus and registration

“Syphilis is a challenge across the US. We're trying to address it by sharing data, being as transparent as we can be, collaborating with people, and being innovative in the way that we share the information. In the past it was very manual to share information. Now with GIS, we can get this information out there quickly so that people can join us in this battle against this disease.” - Diana Ibrahim

Methodology: choices for the Top 10 were filtered to be practical for California providers and directly relevant to San Bernardino County workflows: first, legal and safety screens - state laws like AB 3030 and SB 1120 that require disclosure of generative-AI communications and qualified human review guided exclusion or adaptation of any prompt that could produce standalone clinical messaging (see the Holland & Knight summary on California's healthcare AI laws); second, operational impact - priority was given to use cases that demonstrably free clinical staff from clerical burden, such as AI-driven prior-authorization automation that “frees staff for patient care”; and third, local capacity to deploy and govern tools, using regionally focused training and rollout templates from the San Bernardino County Superintendent's AI resources so teams can assess readiness, build roadmaps, and train reviewers before scaling.

The resulting list favors prompts and workflows that reduce paperwork, protect patient-facing communication, and fit within California's evolving regulatory guardrails.

Selection Criterion	Why It Matters	Source
Regulatory & Safety	Ensure prompts comply with AB 3030 / SB 1120 (disclosure, human review)	Holland & Knight summary of California healthcare AI laws
Operational Impact	Prioritize use cases that cut paperwork and free clinical time	Nucamp AI Essentials for Work bootcamp syllabus (local example)
Local Readiness & Training	Pick prompts deployable with county-level training, policies, and roadmaps	San Bernardino County Superintendent AI resources for educational partners

DAX Copilot (Nuance) brings ambient, voice-powered documentation into the clinician's workflow by drafting specialty-specific notes directly inside Epic - acting as a “copilot” to reduce clerical burden, improve note quality, and let clinicians focus on patients instead of screens.

Built from Nuance's Dragon technology and integrated with Epic (per Epic's announcement), DAX captures conversations via mobile apps like Haiku and generates standardized summaries for review, with turnkey learning guides and best-practice tips available from Microsoft's DAX Copilot resources.

Journalists note the move widens DAX's reach across Epic sites, making adoption more practical for health systems (and for California providers this means smoother EHR integration and centralized workflows).

For teams planning pilots, vendors even outline rapid setups for smaller clinics - so instead of late-night charting, clinicians can spend that reclaimed time in the exam room, improving patient connection and access to care.

Feature	Detail
Epic integration	Native DAX Express integration with Epic for ambient clinical documentation
US availability	Dragon/DAX Copilot general availability in the United States (May 1, 2025)
Typical small-clinic setup	Personal-use integration setup time for small clinics (~3.5 hours)

“Dragon Copilot is a complete transformation of not only those tools, but a whole bunch of tools that don't exist now when we see patients. That's going to make it easier, more efficient, and help us take better quality care of patients.” - Anthony Mazzarelli, MD

Doximity GPT is a clinician-focused generative-AI assistant that sits inside the Doximity ecosystem and quickly drafts the routine paperwork that eats clinician time - referral letters, insurance appeals, patient handouts, and compassionate message replies - so teams can finish admin tasks during the day instead of after-hours; readers can see a hands-on overview at Healthcare Huddle and learn how the broader Doximity network supports HIPAA-safe workflows on Doximity's site.

Built to be HIPAA-compliant for U.S. clinicians, the tool emphasizes human review (every AI draft should be checked before signing) and can be used to generate histories, progress notes, and updated assessment-and-plan drafts that streamline rounds and reduce inbox backlog - clinicians report ending the day with fewer unanswered messages and more time for patients.

For California providers weighing pilots, the key guardrails are vendor BAAs, secure deployment environments, and clear review workflows to keep PHI safe while reclaiming clinician time.

Feature	Clinical Benefit
Doximity GPT hands-on overview at Healthcare Huddle	Saves administrative hours and standardizes correspondence
Doximity official clinician network and app information	Works where clinicians already communicate and call patients
HIPAA-compliant AI deployment guidance from MedCram blog	BAAs, secure environments, and mandatory human review protect PHI

“like waving a magic wand.” - Tina Chu, MD (reflecting on clearing an inbox and serving patients more efficiently)

Ada Health supplies an AI-powered symptom assessment and clinician-curated medical library that can help patients and frontline staff decide next steps when symptoms are ambiguous - especially useful in a sprawling county like San Bernardino where distance and access shape decisions; explore Ada Health AI symptom checker and medical library at Ada Health AI symptom checker and medical library.

Evidence suggests symptom checkers can achieve acceptable usability and triage accuracy in urgent settings (see the JMIR mHealth study), which means these tools can reliably surface red flags rather than replace clinical judgment.

For chest pain specifically, triage frameworks emphasize targeted questions and fast action - risk factors and prompts that detect acute coronary syndrome, pulmonary embolism, or pneumothorax are a checklist for urgency, and clinicians should not delay an ECG when red flags appear (see the Sullivan Group chest pain triage questions and red flags).

Deployed thoughtfully with human review and local workflows, symptom-assessment tools can reduce “should I go to the ER?” uncertainty at 2 a.m., highlight high-risk presentations, and funnel patients toward the right level of care without creating extra false alarms.

Item	Source / Evidence
Ada - clinician‑optimized symptom assessment	Ada Health AI symptom checker and medical library
Usability & triage accuracy	JMIR mHealth usability and triage accuracy study (2022)
Chest pain triage questions & red flags	Sullivan Group chest pain triage questions and red flags

AIDDISON brings enterprise-grade generative AI to medicinal chemistry, helping teams compress hit‑finding and lead optimization into hours instead of years by virtually screening a universe of more than 60 billion chemical targets and proposing synthesizable routes through an integrated Synthia™ retrosynthesis API - features that matter for California labs and biotechs trying to move candidates into the clinic faster and more sustainably.

Built as a cloud-native SaaS, AIDDISON combines de novo molecular design, pharmacophore and shape-based searches, molecular docking, and ML‑driven ADMET screening so chemists can iterate ideas, prioritize drug‑like candidates, and get concrete synthesis recommendations without hopping between tools; see the MilliporeSigma press release on AIDDISON and the product overview at Sigma‑Aldrich for technical detail.

Trials and industry analyses suggest these platforms can shrink discovery time and cost - vendors even cite potential time/cost savings up to ~70% - making AIDDISON a pragmatic partner for San Bernardino research hubs looking to accelerate partnerships between hospitals, regional labs, and startup founders.

Capability	Why it matters	Source
Ultra‑large virtual screening (>60B)	Faster hit identification across vast chemical space	MilliporeSigma press release on AIDDISON drug discovery software
De novo molecular design & ADMET prediction	Generate and rank novel candidates with drug‑like properties	AIDDISON product overview on Sigma‑Aldrich
Retrosynthesis / manufacturability	Bridges virtual design to real‑world synthesis for scale‑ready leads	MilliporeSigma press release on AIDDISON retrosynthesis capabilities

Predictive analytics can turn mountains of EHR data into action for San Bernardino health systems, and platforms in this space (think Merative and similar enterprise tools) are the practical way to do it: literature reviews show predictive models for heart‑failure risk and readmission are an active, maturing field (systematic review of predictive analytics for heart failure and readmission), and real-world deployments demonstrate measurable gains - one integrated heart‑failure model achieved an AUROC of 0.85 (versus LACE ≈0.62) while producing roughly 150 daily risk predictions so care teams can prioritize the right patients each morning.

Those daily, patient‑level scores are most useful when embedded in workflows that family medicine and care managers already run - schedule a follow‑up within seven days, reconcile meds, or assign home‑based services - turning a noisy discharge period into a targeted prevention plan (MultiCare / Health Catalyst heart failure machine learning case study; MGH Institute guidance on leveraging predictive analytics to reduce readmissions).

For county leaders, the bottom line is simple: embed explainable risk scores into daily worklists and team huddles, and the system can shift from reacting to readmissions toward preventing them before the ambulance lights flash.

Metric	Value	Source
AUROC - HF 30‑day readmission model	0.85	MultiCare / Health Catalyst heart failure ML case study
Typical comparator (LACE)	≈0.62	MultiCare / Health Catalyst comparator analysis
Daily risk predictions generated	~150 per day (three‑fold increase)	MultiCare / Health Catalyst deployment metrics

Storyline AI threads behavioral A.I. into telehealth workflows to shift clinics from one-off video calls to continuous, precision care plans that California providers can embed into daily work - helpful where distance and capacity strain access.

The platform pitches measurable gains (4x clinician productivity, 3x more relationship-building interactions) and tackles a common problem - roughly 70% of patients leave visits without fully understanding their plan - by automating tailored education and follow‑up that integrate with care plans rather than just swapping an in‑room visit for a video slot.

Implementation should follow established playbooks: evaluate vendors for HIPAA readiness and EHR integration per the American Medical Association's Telehealth Implementation Playbook, and map staffing, scheduling, consent, and documentation steps as recommended by the U.S. Department of Health & Human Services' telehealth workflow guidance; both resources help meet California's operational and compliance needs.

For San Bernardino clinics, Storyline-style programs offer a vivid payoff - more time at the top of license, fewer patients

“left wondering,”

and telehealth that actually closes the communication loop instead of opening another inbox.

Moxi from Diligent Robotics is a pragmatic, drop‑in teammate for busy California hospitals - rolling quietly down halls to deliver meds, lab samples, PPE, and supplies so nurses can spend more time at the bedside instead of logging steps; hospitals report reclaimed time (nurses can recover up to ~30% of shift time) and measurable throughput gains like reduced walking and faster turnarounds.

Designed for busy, semi‑structured environments, Moxi navigates elevators and doors, carries locked drawers for chain‑of‑custody pharmacy runs, and can be added to new workflows in weeks without major infrastructure changes - see the Moxi product overview at Diligent Robotics for implementation details.

Real deployments include Cedars‑Sinai's twin Moxi pilots (complete with heart‑shaped eyes that make staff smile) and a fleet milestone of 300,000 pharmacy deliveries reported by The Robot Report, evidence that this kind of embodied AI can cut routine friction and lift morale while keeping clinical control in human hands.

Feature	Why it matters	Source
Routine deliveries (meds, lab samples, supplies)	Frees nursing time for patient care	Diligent Robotics Moxi product overview
Proven scale & security	300,000+ pharmacy deliveries; locked drawers for medications	The Robot Report article on 300,000 Moxi pharmacy deliveries
Fast pilots, low infra	Deploy in weeks over existing Wi‑Fi, adaptable workflows	Diligent Robotics Moxi implementation and deployment notes

“We love Moxi… they not only provide an opportunity to improve workflows and be more efficient, but they're a fun thing to see around the halls.” - Melanie Barone, RN (Cedars‑Sinai)

For San Bernardino hospitals, clinics, and senior‑care sites wrestling with long corridors and stretched staffing, Eagle Eye Networks turns existing camera fleets into active safety partners: its cloud Video Management System is camera‑agnostic, scales across multi‑site deployments, and layers AI analytics such as slip‑and‑fall detection, loitering, intrusion, and license‑plate recognition so teams get real‑time alerts and searchable metadata rather than hours of unwatched footage - imagine a system that flags a resident's fall in a dim hallway and pushes an alert to a caregiver's phone before the nurse finishes the next round.

Because analytics can be enabled per camera and run in the cloud, facilities can pilot fall detection or intrusion alerts without ripping out hardware, combine Eagle Eye's smart video search with partner algorithms like NeuraVue for healthcare‑specific alerts, and use analytic event clips to speed clinical review and incident reporting.

For county leaders balancing privacy, cost, and response time, Eagle Eye's cloud VMS and partner ecosystem make proactive monitoring practical across hospitals, long‑term care, and rural clinics in California.

Learn more in Eagle Eye Networks video analytics features overview and the NeuraVue integration with Eagle Eye Networks.

Feature	Benefit	Source
Slip‑and‑fall / fall detection	Real‑time alerts to staff for faster medical response	Eagle Eye Networks video analytics features
Camera‑agnostic cloud VMS	Use existing IP cameras; scale without major hardware changes	NeuraVue integration with Eagle Eye Networks
Per‑camera analytics & smart search	Turn on specific analytics where needed and search footage by metadata	Eagle Eye Networks video analytics features

Verkada's hybrid cloud approach makes duress alerts and access control practical for California health systems by combining on‑device video retention with cloud backup and AI‑driven analytics - so security teams can detect a distress event in real time, lock or unlock doors from the same console, and route verified alerts into incident workflows without swapping out cameras or clogging hospital networks.

The platform's low bandwidth footprint (cameras use roughly 20–50 Kbps while recording) and centralized user management mean multi‑site county networks - from large hospitals to rural clinics - can scale access policies, SSO, and MFA consistently while local staff handle site‑specific responses; see Verkada's hybrid cloud architecture for details.

Ongoing firmware updates and cloud transcoders deliver improved motion search and face/vehicle analytics that speed investigations, and prebuilt APIs let teams pipe duress events into EHRs or safety platforms so alerts become actionable tasks instead of unanalyzed footage - an operational payoff that turns physical security into a clinical safety tool rather than an IT burden.

Learn more about Verkada's camera and AI capabilities for healthcare deployments.

Feature	Why it matters	Source
Hybrid cloud architecture	On‑device storage with optional cloud backup for resilience and selective uploads	Verkada hybrid cloud architecture overview
Minimal bandwidth & scalability	Continuous recording with low upload rates and enterprise bandwidth controls	Verkada hybrid cloud video storage details
AI‑driven analytics & integrations	Real‑time alerts, advanced search, and APIs to integrate duress alerts with IT and clinical systems	Verkada enhanced search and AI features

“We've gone from needing a dedicated technician at each of our 240 locations to just two people managing all 3,000 cameras across these sites.” - Randy Haan, Director of IT Infrastructure, The Salvation Army, Western Territory

For rural clinics in San Bernardino County where upload speeds and IT budgets are tight, Rhombus Systems' edge‑focused cameras make surveillance smart and practical: by running object detection and analytics on the camera itself, the R2 family can cut upload bandwidth by up to 10x and produce alerts 2–3x faster than older approaches, meaning a fall in a dim hallway can trigger a 30‑second clip upload instead of sending hours of empty footage to the cloud - so on‑call staff see actionable events, not noise.

Administrators can tune per‑camera AI, schedule cloud archiving to avoid peak hours, and lean on near‑real‑time LAN streaming when available to keep latency under a second locally; Rhombus' docs explain how selective uploads and activity regions limit unnecessary data flow and make deployments scale without a costly server stack.

For clinics weighing pilots, the Rhombus R2 overview and the Camera Bandwidth guide are practical starting points for building low‑bandwidth, high‑signal safety monitoring that preserves privacy and network headroom.

Metric	Value	Source
Bandwidth reduction (R2 vs R1)	Up to 10×	Rhombus R2 edge AI security cameras blog (bringing AI to the edge)
Typical notification latency (on‑camera AI)	~30 seconds	Rhombus R2 edge AI security cameras blog (notification latency details)
WAN streaming bandwidth (typical)	~120–1,400 Kbps (varies by camera/resolution)	Rhombus camera bandwidth support article (Camera Bandwidth guide)

San Bernardino healthcare leaders can get meaningful wins without a moonshot: choose one high‑impact, low‑risk pilot (think AI‑driven prior‑authorization automation that literally trades an after‑hours charting hour for extra bedside time) and wrap it in simple governance so staff and patients stay protected.

Start by standing up a multidisciplinary AI governance committee, codifying policies and incident procedures, and building role‑based training and auditing into any launch - steps that match the practical checklist outlined by Sheppard Mullin for healthcare AI programs to manage risk while enabling benefit.

Anchor pilots to national guidance so local policy keeps pace (the NAM Artificial Intelligence Code of Conduct is a useful alignment framework) and mind California rules like AB 3030 that require clear disclosure and human‑in‑the‑loop workflows.

Finally, invest in workforce readiness: short, practical courses that teach prompt design, evaluation, and safe deployment help teams move from curiosity to controlled rollout - Nucamp's AI Essentials for Work is one accessible option to get clinicians and staff fluent in those skills before scaling.

Next Step	Why it matters	Source
Form an AI governance committee	Oversight, ethical review, incident response	Sheppard Mullin AI governance program checklist for healthcare
Start a single pilot (e.g., prior‑auth automation)	High impact, lower risk; quick staff time savings	Nucamp AI Essentials for Work - prior-authorization automation pilot example and registration
Train and audit regularly	Role‑based training and monitoring reduce errors and compliance gaps	Nucamp AI Essentials for Work syllabus - training, prompt design, and evaluation

“People are scared of dying, they're scared of losing their mom, they're scared of not being able to parent and walk their child down the aisle. How can we start using the power of these tools, not through a lens of fear and reluctance, but to create a culture change from ‘doctor knows best' or ‘patient knows best' to ‘person powered by AI knows best'?”