Slides
Page 1
Scientific, Technical & Medical Investor Seminar November 3, 2026
Page 2
DISCLAIMER REGARDING FORWARD-LOOKING STATEMENTS This announcement contains forward-looking statements within the meaning of Section 27A of the US Securities Act of 1933, as amended, and Section 21E of the US Securities Exchange Act of 1934, as amended. These statements are subject to risks and uncertainties that could cause actual results or outcomes of RELX PLC (together with its subsidiaries, “RELX”, “we” or “our”) to differ materially from those expressed in any forward- looking statement. We consider any statements that are not historical facts to be “forward-looking statements”. The terms “outlook”, “estimate”, “forecast”, “project”, “plan”, “intend”, “expect”, “should”, “could”, “will”, “believe”, “trends” and similar expressions may indicate a forward-looking statement. Important factors that could cause actual results or outcomes to differ materially from estimates or forecasts contained in the forward- looking statements include, among others: regulatory and other changes regarding the collection or use of personal data; changes in law and legal interpretation affecting our intellectual property rights and internet communications; current and future geopolitical, economic and market conditions; research integrity issues or changes in the payment model for our scientific, technical and medical research products; competitive factors in the industries in which we operate and demand for our products and services; our inability to realise the future anticipated benefits of acquisitions; compromises of our cybersecurity systems or other unauthorised access to our databases; changes in economic cycles, trading relations, communicable disease epidemics or pandemics, severe weather events, natural disasters and terrorism; failure of third parties to whom we have outsourced business activities; significant failure or interruption of our systems; our inability to retain high-quality employees and management or to develop capabilities in emerging technologies at an appropriate pace; changes in tax laws and uncertainty in their application; exchange rate fluctuations; adverse market conditions or downgrades to the credit ratings of our debt; changes in the market values of defined benefit pension scheme assets and in the market related assumptions used to value scheme liabilities; breaches of generally accepted ethical business standards or applicable laws; and other risks referenced from time to time in the filings of RELX PLC with the US Securities and Exchange Commission. You should not place undue reliance on these forward-looking statements, which speak only as of the date of this presentation. Except as may be required by law, we undertake no obligation to publicly update or release any revisions to these forward-looking statements to reflect events or circumstances after the date of this presentation or to reflect the occurrence of unanticipated events. 2
Page 3
Overview and Strategy Kumsal Bayazit CEO 3
Page 4
STM Investor Seminar Agenda Overview and Strategy Kumsal Bayazit, CEO Slides 5-13 Primary Research Gemma Hersh, SVP, Primary Research Slides 14-22 Decision Support Tools overview Kumsal Bayazit, CEO Slides 23-39 • Research Solutions Maxim Khan, SVP Product • Nursing Solutions Paul Crockett, SVP Product Technology Jill Luber, CTO Slides 40-48 Key takeaways Kumsal Bayazit, CEO Slide 49 Q&A 4
Page 5
STM position within RELX H1 2026 RELX revenue RELX adjusted operating profit Overview and Strategy STM 28% STM 30% 5
Page 6
STM revenue by geography and by type H1 2026 Rest of world 36% Europe 22% North America 42% Transactional 22% Subscription 78% TypeGeography 6 Overview and Strategy
Page 7
STM revenue by segment H1 2026 7 Decision Support Tools (Databases, Tools & Electronic Reference) Primary Research Overview and Strategy
Page 8
We operate in long-term global growth industries: Research 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025E 2026F 1.9 2.0 2.1 2.2 2.4 2.5 2.6 2.8 2.9 3.0 3.1 3.3 5% CAGR Global R&D expenditure US $ trillions, constant prices, purchasing power parity Academic & Government Corporate Values are for OECD + Key Countries, forecasts include stimulus additions from planned Government spending in USA, EU and Japan. Sources: OECD (March 2026), UNESCO (Feb 2021; used for historical data only), World Bank (Jul 2026), IMF (Nov 2025). Note: The OECD restates prior figures and forecasts as new information is made available to them; this can cause totals and growth rates to be different to previous reports. 8 Global researchers, millions 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025E 2026F 7.9 8.2 8.5 8.9 9.4 9.8 10.3 10.8 11.4 11.8 12.4 13.05% CAGR • R&D investment drives innovation, economic growth, and solutions to societal challenges • Consistent, long-term growth in number of researchers • Technology advancements continue to improve researcher productivity Consistent long-term growth Overview and Strategy
Page 9
Global practising nurses and physicians, millions We operate in long-term global growth industries: Health 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024E 2025E 2026F 10.1 10.5 10.8 11.0 11.4 12.2 12.5 12.5 12.7 13.3 13.8 14.3 3% CAGR Global health expenditure US $ trillions, constant prices, purchasing power parity 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024E 2025E 2026F 32.5 33.7 35.0 36.5 36.7 38.3 39.9 41.1 42.4 43.8 45.3 47.0 3% CAGR Practising physicians Practising nurses Values are for OECD + Key Countries. Sources: OECD (March 2026), World Bank (Jul 2026), IMF (Nov 2025), Fitch Solutions (BMI download: March 2026). Note: The OECD restates prior figures and forecasts as new information is made available to them; this can cause totals and growth rates to be different to previous reports. 9 • Aging population • Longer and healthier lifespan Consistent long-term growth • Consistent, long-term growth in number of healthcare professionals Overview and Strategy
Page 10
We serve a wide range of customers Research institutions and funders • >10,000 academic and government institutions • >95% of top 1,0001 academic and government institutions • >99% of Nobel Laureates in science since 2000 R&D intensive corporates • >2,000 corporates • >95% of top 100 biopharmaceutical companies2 • ~80% of top 100 R&D driven corporations2 Health • >1,200 medical schools and >3,800 nursing institutions globally • >5,500 healthcare institutions 1. ‘Top’ measured by volume of published research. 2. Top 100 by R&D spend. World2000 and Evaluate Pharma. 10 Research institutions Research funders Life Sciences Biopharma Other R&D intensive corporates Nursing & Medical Schools Researchers Hospitals Overview and Strategy
Page 11
Strategic priorities Expand customer value in Primary Research • Continue to deliver industry-leading research quality, article volume, and user experience at a lower effective price Deliver increasing customer value with our Decision Support tools • Enhance value of existing analytics solutions and expand into new use cases by combining deep domain expertise, leading evidence-based scientific content, advanced analytics and evolving technology Leverage global, modular technology platform to build segment-leading products and drive process innovation • Continue to build technology agnostic and multi-model scalable technology platforms, to accelerate pace of product innovation, organisational speed and agility, and to drive efficiencies 11 Overview and Strategy
Page 12
Underlying revenue growth +2% +2% +2% +2% +2% +1% +3% +4% +4% +4% +5% +6% 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 H1 2026 Underlying revenue growth 2024 and prior growth includes print & print-related 12 • Improving long-term revenue growth trajectory by: – Continued high-quality volume growth in Primary Research – Increasing pace of product innovation in Decision Support Objective: Continue on improving long-term growth trajectory Overview and Strategy
Page 13
Underlying adjusted operating profit growth +3% +3% +3% +2% +3% +1% +3% +5% +4% +5% +7% +8% 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 H1 2026 2024 and prior growth includes print & print-related 13 • Improving long-term profit growth trajectory • Manage cost growth through process innovation, agility and speed Objective: Continue to manage profit growth ahead of revenue growth Underlying adjusted operating profit growth Overview and Strategy
Page 14
Primary Research Gemma Hersh SVP, Primary Research 14
Page 15
We help validate, improve and disseminate science Validate that articles are sound science and a good fit to our journals Help authors improve their articles and edit for accuracy and completeness Drive visibility, usability and global reach on leading platforms • >37,000 editors • ~105,000 editorial board members • >1.9m reviewers • 94% of articles have content changes in peer review • 95% of articles have content changes in production • >3,000 journal brands • >4.6m article submissions1 • ~6 authors per article • 2+ peer reviewers per article • >800,000 articles published1 • 2.9bn articles consumed1,2 • >24m articles archived Publish & disseminate Receive and assess submissions Manage peer review Edit & prepare Archive & promote Industry-leading, scalable journal ecosystem Researchers write up their research and seek the right journal for publishing it. Editors drive the publication process by evaluating submissions, overseeing the review process and providing feedback to authors. Peer reviewers are experts in their research field; they help validate the science and suggest improvements. 1. LTM to June 2026 2. Unique article views and downloads 15 Primary Research
Page 16
Clear leader in quality Article and Citation Share 18% 30% Article Share Citation Share 100% 100% Article and citation share1 Share of articles per quality tier (FWCI2) Share of articles in top 10% journal FWCI tier Share of articles in 10-25% journal FWCI tier Share of articles in bottom 50% journal FWCI tier Share of articles in 25-50% journal FWCI tier Higher quality Lower quality 40% 25% 4% 12% 4%Elsevier Publisher A Publisher B Publisher C Other 1.46 1.11 1.18 0.96 FWCI2 (Industry average = 1) 1. Share of articles per publisher (published in 2021-2024) and share of citations (citations in 2021-24 in relation to articles published in 2021-2024) 2. Field-Weighted Citation Impact (FWCI) for articles published in 2021-2024. FWCI is the ratio of citations received for each article relative to the normalised, expected average of 1.The ratio for each article is normalised based on the expected number of citations by article type, subject field, and publication year Source: Scopus data Primary Research 16
Page 17
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2.6 4.2 +5% CAGR Global industry published articles (millions) Full Year Industry: Global published articles Primary Research 17 • R&D spend growing across government, higher education & business sectors • Number of researchers growing globally • Artificial Intelligence accelerating scientific discovery and improving researcher productivity Total published articles growing globally
Page 18
Elsevier submissions: strong growth Primary Research H1 Full Year 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 H1 2026 1.3 4.2 +12% CAGR >+20% 18 • Breadth of portfolio across disciplines and tiers • Reputation for quality and high-impact journals • Leverage technology to continue to improve user experience for authors, editors, reviewers Strong growth in submissions Elsevier articles submissions (millions)
Page 19
H1 Full Year Elsevier published articles: strong growth Primary Research 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 H1 2026 0.4 0.8 +7% CAGR 19 • Expanding journals portfolio and global presence • Article transfer service finds home for all appropriate articles • Innovation in editorial tools and scalable processes grow output without compromising quality Publishing higher volume of high-quality articles Elsevier published articles (millions) +7%
Page 20
Elsevier revenue per article: improving value Primary Research 20 • Submitted articles growing double digits • Published articles growing mid-to- high single digits • Elsevier Primary Research revenue growing below published articles Increasing value per dollar spent Elsevier submissions, publications, revenue and revenue per article1 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Submitted articles Published articles Primary Research revenue 100 Revenue per article Indexed to 2015 1. All values represent 3-year rolling averages for Primary Research articles (pay-to-read and pay-to-published) and revenue (in USD) Primary Research growth strategy:
Page 21
Creating measurable value for authors, editors and reviewers leveraging sophisticated analytics Keeping human judgement at the core, underpinned by deep domain expertise and a global community of subject matter experts Primary Research 21 Safeguard research integrity and trust • Screening from submission to after publication • Multiple detection signals: AI misuse, authorship, citations, paper mills and data fabrication • Detection to resolution, with traceable decisions • An arms race, as with fraud in other industries, leveraging Risk capabilities and know-how Live and scaling In testing • Streamlined submission process: faster, better-quality submission • Online peer review process: higher transparency, less friction and faster decisions • High author, reviewer and editor satisfaction • Improved speed and acceptance rate • Change management at scale • AI editor tools to streamline the work and remove admin • Agentic content enrichment process decreasing time from acceptance to publication • Rolling out LeapSpace capabilities e.g. Claim Radar Higher quality, faster process and better user experience
Page 22
Primary Research: key takeaways • Segment growth continues to be driven by volume growth • Elsevier’s published articles growing faster than the industry • We continue to increase customer value • Our objective in Primary Research is to deliver continued gradual improvements in growth 22 Primary Research
Page 23
Decision Support Tools (Databases, Tools & Electronic Reference) 23 Kumsal Bayazit CEO
Page 24
We help customers solve complex use cases Product Category Buyers Example use cases Research Solutions Nursing Solutions • Universities and research institutes • Funding bodies and government agencies • R&D-intensive corporations • Discover, synthesise and utilise scientific content • Identify research topics, find collaborators and funding • Write and publish manuscripts • Support R&D for drugs, materials and new product development • Support regulatory approval and compliance • Nursing schools • Hospitals and health systems • Student: build knowledge, apply skills, and assess exam readiness • Faculty: create course and assess student understanding • Practitioner: Support ongoing professional development and point of care • Medical schools • Hospitals and health systems • Pharmacies • Corporations • Student: build knowledge, apply skills, and assess exam readiness • Faculty: create course and assess student understanding • Physician: Support professional development and point of care • Patient: Education and engagement Decision Support Users • Grant officers • Academic leaders • Nursing students • Nursing faculty • Practising nurses • Medical students and faculty • Physicians, pharmacists and patients • Clinical operations teams Expand our suite of high value solutions: Improve existing use cases, expand into new use cases, integrate where customers work Conduct & Publish ResearchManage & Evaluate Research Medical Solutions and Other • Set funding priorities • Manage funding awards and assess outcomes • Build research teams • Manage faculty recruitment / progression • Universities and research institutes • Funding bodies and government agencies • R&D-intensive corporations • Academic researchers and students • Corporate scientists and engineers 24
Page 25
Four drivers of higher customer value improving revenue growth trajectory • Increasing value of our existing high value tools • Launching new high value products • Building interoperable product suites • Integrating into customer environments Decision Support The combination of one or more of these drivers improves customer value 25
Page 26
Four key capabilities 1. Deep customer understanding and domain expertise 2. Leading content and data 3. Sophisticated linking and analytics • Derived data in knowledge graphs with billions of linked data points • Licenced and curated databases, rights- cleared and comprehensive • Domain depth, e.g. engineering, chemistry, life-sciences • Semantic enrichment through taxonomies and ontologies for scientific reasoning • Expert community of editors, editorial boards, peer reviewers and advisory boards • Deep customer relationships • Extensive industry experience • Proprietary scientific journal and medical portfolio of highest quality • Decades of history, expert-curated and continuously updated • Technology agnostic and multi-model • Private and secure, customer data not used to train frontier models • Global, modular, flexible architecture: build once, reuse everywhere 4. Advanced technology in global platforms • Human oversight across curation, integrity checks and evaluation of AI outputs • Multimedia health content across continuum of care • Evidence-based verifiable AI outputs, transparent by design, consistent and traceable Decision Support 26
Page 27
Research Solutions 27 Maxim Khan Senior Vice President, Product
Page 28
Who What Key Elsevier products Illustrative research journey Academic / Corporate ResearchersResearch offices and academic leaders Help researchers discover and access content, identify topics and funding opportunities, find collaborators and select appropriate journal for submissions Help funders and policy- makers set funding priorities, manage awards and assess funding impact • Scopus • ScienceDirect • Reaxys • Engineering Village • Embase • Scopus • Data Monitor • Funding Institutional • Pure • Interfolio • Digital Commons • Analytical services • SciVal • Mendeley • Editorial Manager • SSRN • Knovel • SciBite Help academic leaders shape research strategy, build research teams, assess funding outcomes and impact versus peers, and manage faculty recruitment and progression Funders / government Define and execute funding strategy Conduct research Define and execute research strategy Manage faculty & research processes Publish research Help academic researchers publish research articles, preprints and datasets, conduct peer reviews, and perform editorial duties Support all stakeholders across the research ecosystem LeapSpace Decision Support Manage & Evaluate Research Conduct & Publish Research Apply research Help corporate researchers with drug and new product development 28 • PharmaPendium
Page 29
LeapSpace: Integrated AI-enabled research workspace 7k+ publishers1 500k+ books >55% full-text coverage of citations2 2.6bn cited references1 109m+ records1 • Launched in January 2026 as an evolution from ScienceDirect AI launched in March 2025; continued rapid expansion of capabilities since launch Key value drivers: • Improves outcomes for tasks across the full research journey such as ideation, exploration, planning and collaboration in one seamless workspace • Grounded in comprehensive trusted content including Scopus, Elsevier’s proprietary full text content, licensed content from select publishers, funding data and user data • Evidence-based and verifiable outputs which support creativity, curiosity, and critical thinking through transparency and explainability with Trust Cards • Agentic capabilities including deep research, Claim Radar, Writing Coach and funding database Content and algorithm standards are guided by the LeapSpace Advisory Board, an independent, external group of international academics World Class Data ~25M full-text articles2 1. Scopus data coverage, 1.9K publishers licensed for full-text 2. LeapSpace draws on ~25m full-text covering approximately 56% of Scopus last-five-years citation share Decision Support 29
Page 30
LeapSpace demo Decision Support 30
Page 31
LeapSpace: Very positive customer feedback and rapid usage growth • Very positive customer feedback on value: Better outcomes, greater efficiency, trust and control • Go-to-market approach: Subscriptions to Academic and Government institutions and Corporate customers • Rapid usage growth: Growth in activated users, monthly active users 31 Decision Support
Page 32
Nursing Solutions 32 Paul Crockett, Senior Vice President, Product
Page 33
Who What Key Elsevier products Aspiring Healthcare Professional Illustrative Nursing Journey Support the nursing journey from education to practice Decision Support Clinical Practice Educators and Nursing Students Help educators facilitate student learning and mastery; help students build foundational knowledge Help educators assess student licensure exam readiness; help students prepare for licensure exam success Help educators develop student clinical readiness; help students apply clinical judgement skills • Sherpath AI • ClinicalKey Student • Osmosis • Complete Anatomy • Shadow Health • HESI Build Knowledge Apply and Practise Assess and Review ∞Non-linear, continuous learning cycle across Build Knowledge, Apply and Practise, Assess and Review and Clinical Practice • ClinicalKey Nursing AI • Care Planning • Clinical Skills Help hospitals transition new graduates into practice; help practising nurses deliver evidence-based care and with continuing education Nursing Practitioners 33
Page 34
Nursing education: Higher customer value through connected enterprise offering Predictive analytics Student Programme Institution Dashboards show faculty and deans each student's likelihood of passing the required NCLEX1 exam, so they can intervene early AI-enabled teaching Build knowledge: Sherpath AI Conversational tutor based on Elsevier trusted content; provides answers and builds lesson and study plans Apply Clinical Judgment: Shadow Health Voice-driven patient simulations using speech-to-text and language processing Assess & Review: HESI Personalised learning plans and practice tests that remediate each learner's gaps Modular offering can be connected to provide opportunity for end-to-end learning support 1. National Council Licensure Examination - the standardised, computer-adaptive licensure exam required to obtain a nursing licence in the United States and Canada + our education solutions integrate with customers’ learning management systems + 500k Elsevier Adaptive Quizzing Questions 100k Elsevier HESI Assessment Questions 15k Sherpath lessons Trusted content Curated healthcare education content continually updated by experienced nurse practitioners and educators 2k+ Whiteboard-style learning videos, patient simulation models and assignments Decision Support + 34
Page 35
Nursing education: Enabling positive student outcomes Build Knowledge Assess & Review Apply Clinical Judgment Students who used Sherpath and Elsevier Adaptive Quizzing scored 15% higher on practice exams than students who did not1 98% of students scoring the HESI benchmark of 850+ on their HESI practice exam, pass their licensure exam on their first attempt3 74% of students who use Shadow Health simulations increase their clinical judgment skills (explicitly measured on the licensure exam)2 Decision Support 1. Study of c.600 students. 2. Study of c.2,000 students. 3. Study of c.1,750 students 35
Page 36
Nursing education: Sherpath AI is our core AI-enabled interactive learning assistant Trusted Content Verifiable Citations Multimedia Content Interactive Dialogue Personalised Context Responsible AI Independent learning with trustworthy resources Anytime/Anywhere Q&A Expands faculty bandwidth Personalised Content Generation Launched in Oct 2024 Combines Elsevier's trusted content with advanced generative AI to support students and educators Sherpath AI Decision Support 36
Page 37
Nursing education: Sherpath AI Demo Decision Support 37
Page 38
2025 revenue $0.2M Limited purchasing of Elsevier nursing products at the course level No Elsevier contract: school purchasing on a seminar-by-seminar basis 2027F revenue $1.7M Purchasing the full suite of Elsevier nursing flagship products at the programme level Multi-year Elsevier contract: school decided to sign a three- year contract Greater customer value through expanded relationship Nursing education: Enterprise level offering driving customer value expansion Spend per student more than doubles in our highest package tier Nursing Overall Practical Nursing Associate Degree Bachelor’s Degree 2.2x 2.2x 2.1x 2.5x Case study: A leading US university Successful user outcomes from using a combination of our solutions drive greater customer value and penetration of our nursing courses Decision Support 38
Page 39
Decision Support: key takeaways • Growth in global research drives increased demand for trusted evidence-based tools • We continue to add greater value in our Decision Support tools by improving existing use cases, expanding into new use cases, and embedding where customers work • Our objective in Decision Support is to continue to translate higher customer value into continued gradual improvements in growth with our increasing pace of product innovation • Innovation is enabled through the combination of our deep customer and domain expertise, leading content and data, sophisticated linking and analytics, and advanced technology platforms 39 Decision Support
Page 40
Technology Jill Luber Chief Technology Officer 40
Page 41
RELX technology Build new products and support innovation Build and improve our big data, analytics, and technology platforms Ensure platforms are increasingly reliable, scalable, secure, and compliant Support operational efficiencies via automation and re-tooling • $2bn RELX annual technology spend • More than 12,000 technologists at RELX • 15+ years of experience with big data and AI/ML • Technology agnostic • Leverage approaches across RELX Global Scale 41 Technology
Page 42
STM Technology: Extracting context and insights at scale • >109m publication records, from >48k sources; >179m patents; >3m preprints; >9m grants; c.8m policy doc links • >54k drug database records; clinical trial data; clinical guidelines • Entity extraction, tagging and linking to structure unstructured content at scale • Expert-informed taxonomies and ontologies to normalise, disambiguate and contextualise entities • Semantic relationships across disciplines and domains to connect concepts, evidence and knowledge • Specialised agents for discovery, synthesis, analysis and decision support • Agent orchestration to coordinate tasks, tools and knowledge across workflows • Human-guided intelligence with validation, transparency and feedback built in SOURCE CONTENT DERIVED DATA CONNECTED CONTEXT Structured and unstructured content Connected context, grounded in domain expertise Intelligence & Agents AI HUMAN EXPERTISE, GOVERNANCE & OVERSIGHT AI SUPPORT FOR CURATION, VALIDATION & ENHANCEMENTS (continuous feedback loops) STM product examples Clinical trial data Policy documents Drug databases Journals, books Patents INTELLIGENCE > > > Structure, enrich and validate content into trusted entities Connect entities and relationships with full semantic context Reason over connected context plus user context to generate insights and actions + user context 72M 11M 433M 54K 83K 76M 9M 374M 51K Technology 42
Page 43
AI outputs that are more accurate, relevant and verifiable Content, data and analytics to deliver better outcomes Technology Definition Value Our strengths Leading scientific and medical content and data Elsevier published content: journals, society guidelines, clinical overviews, monographs, multimedia content Proprietary Licensed and curated 3rd party publisher content, open access, patents, grants, policy, drug information, preprints High quality; proprietary Breadth and depth: 300 disciplines, domain depth • Editorial and production integrity • Peer review and domain network • Research integrity • Thousands of different sources • Curated • Rights management, version of record Sophisticated linking and analytics Enriched data and knowledge graphs: citations, authors, diseases, chemical reactions, substances, drugs & symptoms Derived data Semantic layer Domain-specific taxonomies, millions of concepts across research, biology, chemistry and health Analytic layer: Structured, connected, normalised and disambiguated data Semantic layer: Domain- specific scientific concepts understanding • Billions of data points • Longitudinal data: expert-curated and continuously updated • Structured networks and knowledge graphs • 50+ domain taxonomies • Millions of concepts • Deep data science and domain expertise Connected context Semantic relationships across disciplines, connecting concepts, evidence and knowledge Connected context to generate domain specific insights and actions • Multi-model architecture, technology agnostic and built to absorb change • Human validation, transparency and feedback built in Context layer 43
Page 44
Reaxys PharmaPendium Embase Resolve Align identities, synonyms, registry numbers and nomenclature Retrieve Pull relevant chemistry, PK, safety and regulatory evidence Harmonise Normalise units, parameters, ontologies and study context Deliver Return structured, analysis-ready content and data AI-ready surface Elsevier AI-enabled solutions Customer environments Reliable, dependable evidence-based layer Consistent and repeatable output: one scientific task, delivered the same way every time 44 Example: chemistry and biomedicine Technology Curated content & data
Page 45
Combination of deep data science and domain expertise to optimise outputs Example Customer question: Output: ‘Does gene X regulate protein Y expression in human cell lines?’ STAGE 1 Understand • Intent What the user actually wants • Concepts Named entities and relations • Constraints Time, domain, language STAGE 2 Retrieve Lexical exact terms Semantic meaning vectors Structured citation graph, ontology ACROSS Text Abstracts Tables FiguresEquations STAGE 3 Evidence-based, repeatable answer • Citations claim → source mapping • Explainable why this evidence won • Repeatable same query, same answer Technology 45
Page 46
Technology agnostic, multi-model approaches drive quality, speed, and cost Technology Example: Model selection process BENCHMARK Measure model quality • Suppliers Anthropic, Google, OpenAI, Mistral • Families Opus, Sonnet, GPT, Pro, Flash • Hosting AWS, Azure, Direct • Metrics quality, latency and cost quality gaps OPTIMISE Reuse & evolve configurations • Optimisation axes quality, throughput, cost • Spec-driven configurations generated per model family • Reinforcement continuous improvement from feedback • Version history traceable, reversible, auditable optimised setup MONITOR Guard production quality • Observability dashboards, variance alerts • Traffic routing by customer, product, outage • Drift response re-tune on drift signal • Regression response failover to backup model drift and regressions feed the next benchmark 46
Page 47
Process innovation and agility • Spec-driven development enhances code quality and shortens development lifecycles • Reusable, modular platforms shorten launch times across products Continuous process innovation, now accelerated by AI, to manage cost growth below revenue growth Technology • Automated, reusable content pipelines process growing volumes at lower cost • AI-assisted evaluation of product outputs ensures accuracy and qualityOperations • AI agents improve response speed and consistency • Automated case routing and prioritisation improve operational efficiencyCustomer Support • Technology-assisted analysis of customer conversations surfaces insights • Channel and audience analysis identifies optimal approach to connect with customersSales & Marketing Select examples 47 Technology
Page 48
Summary: Elsevier’s technology platform enables agile innovation at scale Faster innovation and development Build once, reuse everywhere - new AI features ship in days, not quarters Evidence-based answers Precision AI, trust cards and evaluation keep outputs more accurate, traceable, and consistent Cost efficiency AI-assisted development, pre- computed content and prompt optimisation reduce the cost to build, run and support our products Technology flexibility Multi-Model approach - no lock-in; we switch to the most appropriate models as the market evolves Technology 48
Page 49
Key takeaways 49 • We operate in large, global segments with attractive structural long-term growth drivers • We have segment-leading positions helping our customers with complex and critical use cases • We deliver customer value through the combination of domain expertise, leading scientific content, sophisticated linking and analytics and advanced technology platforms • Our objectives are to continue to deliver higher customer value, continue on our improving long- term revenue growth trajectory, and manage profit growth above revenue growth
Page 50
Q&A 50