AI in Healthcare & Pharma Summit schedule

• Why most 2023–2024 pilots stalled at scale, and what separated the organisations that broke through
• The “clean sheet” mindset: asking whether a process should exist at all before applying AI
• Treating AI fluency as foundational infrastructure rather than a bolt-on capability
• What a credible, board-ready scaling roadmap looks like over the next 12–18 months

Cheng Yang, Digital Diagnostics Lead - Tekeda

• Where autonomous, think-plan-act agents are already delivering measurable time savings
• Embedding agents into documentation, prior authorisation, and revenue cycle operations
• Reclaiming clinician hours and easing the documentation burden behind burnout
• Oversight models that keep humans accountable for agent-driven decisions

 Preeti Pabreja, AVP, Software Development - CareCentrix  

• Why fragmented EHR ecosystems remain the single biggest barrier to AI at scale
• Building the data foundations before scaling advanced AI use cases
• The CMS Interoperability & Prior Authorization rule as a 2026 compliance catalyst
• Standardising and governing data sharing across providers, payers, and pharma

• Using machine learning to separate urgent signals from clinical noise
• Reducing cognitive overload and improving situational awareness at the bedside
• Lessons from ICU decision support and sepsis surveillance deployments
• Measuring clinical trust and adoption beyond the proof-of-concept

• A practical, validation-first approach to selecting clinical AI tools
• Embedding LLM-powered tools directly and safely into EHR workflows
• Guardrails for hallucination, accuracy, and specialty coverage
• Evidencing real-world clinical performance, not benchmark scores

Anemone Kasasbeh,Data Science Manager– Ascension

Amir Tahmasebi, Head of AI Algorithms and Infrastructure – BD

• Shifting from reactive treatment to anticipatory, data-driven care
• Identifying at-risk populations earlier across diverse communities
• Designing for equity so models serve underserved groups fairly
• Connecting predictive insight to action across provider and payer workflows

 Anemone Kasasbeh, Manager of Data Science and Clinical AI - Ascension  

• Planning, executing, and monitoring trials with less linear human intervention
• AI-driven protocol simulation and real-time enrolment optimisation
• Realistic gains in study startup timelines and database lock
• What operational readiness for autonomous trial ops actually requires

• Where virtual control arms are delivering credible efficiency gains today
• High-data therapeutic areas where twins work, and where they don’t
• How evolving regulatory guidance is shaping confidence in adoption
• Pairing operator experience with technology for a commercial-free view

• Breaking silos and building cross-functional alignment at scale
• Embedding governance without slowing scientific progress
• Moving from efficiency gains to genuinely new decision-making capability
• Translating enterprise AI investment into measurable patient outcomes

Manohar Boorlu, Associate Director, AI Lead –Novartis

Lily Xu, PhD, Director, Clinical Data Science – Vertex Pharmaceuticals

• Why most AI programmes stall on people problems, not product problems
• Managing cognitive load, alert fatigue, and change resistance
• Building AI fluency across clinical, operational, and research roles
• Sustaining trust and momentum after the initial pilot wins

• The emergence of AI that reasons about the physical world the way LLMs reason about language
• Implications for preclinical molecule design and drug delivery
• Manufacturing yield and materials science as new AI frontiers
• Positioning R&D strategy ahead of an accelerating curve

• • Moving governance from a compliance checkbox to a genuine strategic enabler
  • Cutting through the noise: separating real capability from vendor hype
  • Explainability, bias auditing, and audit trails built into the model lifecycle
  • Aligning internal playbooks with evolving FDA, EMA, and EU AI Act expectations
    

  Manasi Ghogare, Senior Technical Product Manager, Digital Health – Takeda

  Bickkie Solomon, Director of Pharmacy; PGY-2 HSPAL Residency Program Director – HCA Florida North Florida Hospital

• Quantifying and communicating AI’s return to the C-suite and board
• Building scalable infrastructure that meets regulatory expectations
• Navigating budget constraints and cross-functional resistance
• Defining the metrics that signal an AI programme is genuinely working

Parisa Farzam, Blavatnik Fellow in Healthcare and Life Sciences - Harvard Business School

• Separating what’s real, what’s hype, and what’s worth preparing for now
• Whether multimodal models and digital twins deliver value today
• Infrastructure and ethical considerations for scaling them responsibly
• How leaders should prioritise AI R&D over the next three years

Why model deployment is the beginning of the risk cycle, not the end

• Detecting drift, performance degradation, and unintended consequences in live environments

• Establishing ownership across data science, clinical, regulatory, technology, and business teams

• Designing monitoring systems that protect outcomes without slowing innovation

How AI is changing the division of work across clinicians, researchers, operations, and commercial teams

• Deciding which activities should be automated, augmented, or kept fully human

• Reskilling experienced teams without creating parallel AI organisations

• Building clear accountability when decisions are shared between people and intelligent systems

Moderator: Rafael Areses, Medical Doctor MD DDS OS, Senior Advisor, Author and AI Tech Speaker

Bickkie Solomon, Director of Pharmacy; PGY-2 HSPAL Residency Program Director – HCA Florida North Florida Hospital

 Hospital

Bringing model performance, data quality, bias, and operational risk into one monitoring layer

• Identifying silent failures across changing populations, workflows, and clinical settings

• Creating escalation thresholds and evidence trails for high-risk AI systems

• Turning observability into an operational capability rather than a technical dashboard

• Understanding new attack surfaces created by third-party models, APIs, and training data

• Protecting sensitive health and research data from leakage, manipulation, and misuse

• Assessing vendor security without relying solely on contractual assurances

• Building resilience against model poisoning, prompt injection, and adversarial behaviour

• • Identifying where AI can remove administrative friction without fragmenting care

  • Redesigning workflows so insight arrives at the right moment rather than creating another alert

  • Preserving professional judgement as recommendations become more automated

  • Measuring impact through capacity, care quality, staff experience, and patient trust

• • Integrating remote monitoring, wearable, and patient-reported data into clinical workflows

  • Separating meaningful signals from continuous streams of low-value information

  • Designing escalation pathways that lead to timely human intervention

  • Demonstrating improved outcomes without increasing pressure on clinical teams

 Erlyn Macarayan, PhD, Vice President of Data Science - PatientsLikeMe (Fuze Health)  

• Clarifying responsibility between providers, vendors, clinicians, and technology partners

• Managing liability when models influence—but do not independently make—decisions

• Establishing processes for incident response, remediation, and transparent communication

• Building partnership agreements that remain workable as technology and regulation evolve

 Amir Tahmasebi, Head of AI Algorithms and Infrastructure – BD  

Integrating regulatory strategy into AI development from the earliest design stages

• Defining validation standards for models that continue learning or changing after deployment

• Creating traceable evidence across data provenance, performance, and human oversight

• Preparing globally scalable approaches across FDA, EMA, and emerging AI frameworks

Integrating regulatory strategy into AI development from the earliest design stages

• Defining validation standards for models that continue learning or changing after deployment

• Creating traceable evidence across data provenance, performance, and human oversight

• Preparing globally scalable approaches across FDA, EMA, and emerging AI frameworks

 Cheng Yang, Digital Diagnostics Lead - Tekeda  

• Choosing where proprietary capability creates genuine strategic advantage

• Evaluating startups, technology providers, research institutions, and platform partners

• Structuring partnerships around shared outcomes rather than isolated pilots

• Protecting intellectual property while enabling data and capability exchange

Manohar Boorlu, Associate Director, AI Lead -Novartis

• Moving beyond broad principles to visible evidence of safety, fairness, and accountability

• Communicating how AI is used without overwhelming patients or frontline professionals

• Responding transparently when models underperform or produce unintended outcomes

• Making trust a measurable operational asset rather than a communications exercise

The 2030 AI Operating Model — What Leaders Must Build Next

• Which capabilities healthcare and pharma organisations should own internally

• How operating models must evolve as AI becomes embedded across every function

• Preparing for increasingly autonomous systems without losing strategic or human control

• The decisions leaders must make now to create durable advantage over the next five years

 Moderator: Rafael Areses, Medical Doctor MD DDS OS, Senior Advisor, Author and AI Tech Speaker