Transforming Clinical Insight Through Local AI Compute: The WWT and Intel Collaboration

In healthcare, data fragmentation has long been one of the biggest barriers to effective, timely care. Physicians often face a complex web of electronic records, imaging archives, and specialist notes scattered across different systems. The result is a familiar frustration: too much data, too many silos, and too little clarity.

To address this, World Wide Technology (WWT) and Intel partnered to explore how local AI compute can turn scattered information into usable clinical insight—fast, secure, and right at the point of care.

The problem: The "data tax" on clinicians

During a recent collaboration between WWT and Intel's Healthcare and Life Sciences AI teams, Dr. Eric Quiñones described the challenge clearly:

The time lost navigating disconnected systems isn't just an efficiency issue; it's a burnout issue. The Pre-Visit Planning Tool (PVPT) was designed to address this by automatically curating and summarizing patient data into a single, easy-to-read timeline, thereby significantly improving practitioners' quality of life.

The solution: Intelligence at the edge

The PVPT demo showcased a paradigm shift in medical data processing. Built with Streamlit and powered by a Llama 3 (3.2B) model, the tool is specifically optimized through Intel® OpenVINO™ to run locally on an Intel AI PC.

As WWT's Ina Johnson demonstrated, the tool condenses what would normally be 15 minutes of manual record review into just 15 seconds of automated summarization.

Key technical advantages:

• Speed: Processes raw longitudinal data and generates summaries in seconds.
• Privacy: Because the LLM runs entirely on the Intel AI PC, no data ever leaves the device. There is no reliance on external APIs or internet connectivity.
• Security: Local inference keeps sensitive patient data secure and within the hospital's regulatory perimeter.

Intel's role: Powering regulated environments

The collaboration with Intel's Life Sciences team demonstrates the power of edge AI in environments where privacy and performance are non-negotiable.

Intel hardware doesn't just accelerate the AI; it makes it practical for daily use. Andrew Lamkin, from Intel, highlighted how this proof-of-concept bridges multiple healthcare pain points:

By using Intel-optimized hardware, clinical AI tools can operate independently of the cloud, remain cost-effective, and be within reach of any provider's existing infrastructure.

The impact on patient care

The benefits extend beyond the clinician's desk. From a patient's perspective, the burden of care coordination is a major stressor. 

Patient advocate Susie Capozza shared: 

By providing a unified chart review, the PVPT empowers clinicians with greater visibility while improving the patient experience, so patients no longer feel they are starting from scratch at every appointment. Additionally, as mentioned, adding treatment and health maintenance protocols to this demo can identify care gaps at the point of care, further improving the clinician and patient experience. 

A path to smarter integration

The PVPT demonstrates that meaningful AI in healthcare doesn't require massive infrastructure or off-site computing. It shows how local, Intel GPU-accelerated systems can streamline care preparation while preserving patient privacy.

For clinicians, it's a time-saver. For patients, it provides reassurance. For hospitals, it's a path toward smarter data integration. For Intel and WWT, it's a tangible example of how thoughtful design can make healthcare more efficient, one curated care journey at a time.