AI, Genomics, and the Israeli Paradox: What Healthcare Leaders Can Use Today
At the 32nd Israel Business Conference, Prof. Yitshak Kreiss, CEO of Sheba Medical Center at Tel Hashomer, laid out a plan that's hard to ignore. Sheba is partnering with Nvidia to push toward decoding most of the human genome using artificial intelligence-an effort he called a "little project" with the potential to map 98% of DNA.
For healthcare teams, this isn't just a tech headline. It's a playbook for how to build capability under constraints, accelerate discovery, and redesign care pathways with measurable outcomes.
Sheba × Nvidia: A practical path to decoding DNA
Genome mapping is still costly and compute-heavy. That's the gap Sheba is closing with Nvidia-combining data, expertise, and infrastructure to reduce time, cost, and error in variant detection and interpretation. The upside is clear: earlier prevention, faster diagnosis, and smarter drug development.
Context matters here. Sequencing costs have fallen but remain significant at scale, and end-to-end interpretation still consumes resources. See the NIH's overview of genome sequencing costs for perspective: NIH/NHGRI factsheet. For the compute and tooling side, Nvidia outlines its healthcare stack here: NVIDIA Healthcare & Life Sciences.
The lesson: don't wait for a perfect budget. Start with focused partnerships, share risk, and define narrow, high-impact use cases where data quality and clinical validation are achievable.
- Data discipline first: standardize pipelines, consent flows, de-identification, and governance. Decide upfront what gets shared, how, and with whom.
- Right-size compute: secure GPU access (cloud or on-prem), containerized workflows, and a clear path for model validation and updates.
- Start where signal is strong: rare disease variant triage, oncology panels, or imaging-genomics integration-paired with prospective clinical evaluation.
- Measure what matters: time-to-diagnosis, cost per interpreted genome, avoidable admissions, and treatment modifications driven by findings.
Changing a conservative profession without breaking trust
Kreiss argues that healthcare hasn't fundamentally changed in centuries-and that now is the moment. Change will stick only if clinicians feel supported, not replaced.
Build confidence with rigorous validation, transparent limitations, and workflow-native tools that shorten steps, reduce clicks, and keep clinical judgment in the loop.
- External validation on diverse cohorts, continual bias checks, and drift monitoring.
- Human sign-off for critical decisions, with clear exception handling and escalation.
- Plain-language model cards and patient-facing explanations of how AI informs care.
The Israeli paradox: crisis fueling growth
Israel is facing one of its hardest periods and still finding ways to grow. Kreiss calls this a mix of enterprise culture, concept-breaking, and a willingness to turn systems once seen as fiscal burdens into engines of economic value.
He points to healthcare as a key example: world-class data, thousands of doctors and researchers, advanced infrastructure, and a startup mindset. Sheba's ARC Network climbed from $1.6B to $5.8B in value since October 7-a signal of how healthcare innovation can expand even under pressure.
What hospital systems can do now
Your advantage is closer than it looks: data, clinicians, research units, and patients who want better answers. Combine those into pilots, partnerships, and clinical trials that create both outcomes and revenue.
- Create an innovation arm or venture studio tied to operations, not just research.
- Use secure data enclaves and consider federated learning to collaborate without moving raw data.
- Partner with global tech and pharma for compute, tooling, and commercialization-on your terms.
Trauma care signals: 99%+ hospital survival
As a reserve brigadier general and former IDF chief medical officer, Kreiss highlighted a striking figure: if patients reach an Israeli hospital alive-soldier or civilian-the chance of leaving alive is above 99%. That improvement is tied to better evacuation, new tech, and disciplined systems thinking.
Translate that to your setting: shorten the prehospital-to-ED interval, align protocols across teams, and practice incident command before you need it.
- Standardized massive transfusion protocols and prehospital blood availability.
- Real-time bed capacity dashboards and OR readiness alerts.
- Joint EMS-ED simulation drills with fast after-action loops.
- Portable imaging and POCUS at the point of care; use approved AI tools where they reduce misses and time.
Bottom line
Healthcare can be both a public good and a growth engine. The path is practical: targeted partnerships, disciplined data practices, and tools that make clinicians faster and more precise. Crisis can accelerate change-but the systems that win are the ones that prepare before the next surge.
If your team is building AI skills for clinical and operational use, these structured resources can help: Courses by job.
Your membership also unlocks:
