Executive Summary
For as long as the pharmaceutical industry has practiced scientific exchange, the model has been the same: evidence is generated, Medical Affairs interprets it, Marketing distributes it, and healthcare professionals search for it. That model is ending. A new layer — large language models, AI assistants, and intelligent search systems — now sits between evidence and the clinical decisions it is meant to inform. Physicians are not replacing their judgment with AI; they are routing an increasing share of their information-gathering through it.
The scale of this shift is no longer speculative. Independent 2026 surveys put generative AI use among physicians at 81–92%, more than double 2023 levels, with no specialty adopting at below 85%. A single clinical-AI platform, OpenEvidence, is now used daily by more than 40% of practicing U.S. physicians and fields over 20 million clinical queries a month. Physicians report that AI meaningfully shapes their clinical decision-making, not just their search behavior. For an industry whose entire commercial and medical model rests on being present at the moment a clinical decision is made, this is not a channel shift. It is a redefinition of what "reaching the HCP" means.
This whitepaper sets out the case, grounded in the uploaded strategic thinking behind travalcon's Future of Scientific Engagement concept and extended with current market research, that the winners of this transition will not be the organizations that produce the most content. They will be the organizations whose evidence is easiest for both a physician and an AI system to find, trust, and correctly apply. That requires five capabilities working as one system — the same five capabilities travalcon has already built as independent frameworks: governed evidence (BCB), modular content (Modular Content Framework™), machine-executable metadata (Tagging & Taxonomy Framework™), a connected semantic layer (Knowledge Graph Framework™), and citation-grade visibility (AI Visibility Optimization Framework™) — now shown as one coherent response to a single market shift.
- 81–92% of physicians now use generative AI in clinical practice — more than double 2023 adoption, no specialty below 85%
- OpenEvidence alone: 40%+ of U.S. physicians daily, 20M+ monthly clinical queries, $12B valuation as of January 2026
- 73% of physicians say AI significantly (12%) or somewhat (61%) influences their clinical decision-making
- AI-in-medical-writing market ~$974M in 2026; real-world evidence solutions market $2.6–3.6B in 2026
- Physicians' top AI tools by reported use: ChatGPT 83%, Gemini 50%, OpenEvidence 36%, UpToDate AI search 32%
The organizations that prepare today will define how healthcare knowledge is accessed tomorrow. This whitepaper explains why, and sets out what preparing actually requires.
1. The AI-Powered HCP Has Arrived
Healthcare professionals face an unprecedented information burden. Medical knowledge continues to double faster than clinical workloads can absorb it, and large language models have emerged as a new layer between information and decision-making — helping clinicians summarize evidence, search medical literature, review treatment guidelines, draft documentation, prepare patient communication, and support clinical education. The result is not physicians outsourcing judgment. It is physicians augmenting it.
1.1 What the 2026 Data Actually Shows
| Metric | 2026 Figure |
|---|---|
| Physicians using generative AI in clinical practice (US AMA survey) | 81%, more than double the 38% recorded in 2023 |
| Physicians using generative AI, cross-market survey (US/UK/Canada/China/Germany/France/Italy) | 92%, no specialty below 85% |
| Physicians expecting their AI use to increase over the next year | 85% |
| Physicians satisfied with the AI-generated answers they receive | 80% |
| Physicians reporting AI significantly or somewhat influences clinical decisions | 73% (12% significantly, 61% somewhat) |
1.2 What HCPs Actually Use AI For
The reported use cases cluster around four needs: medical information retrieval (understanding new studies, comparing treatment options, summarizing publications, exploring unfamiliar disease areas); clinical education (explaining complex concepts, generating educational summaries, board-exam preparation); documentation support (drafting notes, summarizing encounters, reducing administrative burden); and point-of-care knowledge support (reviewing treatment pathways, checking guideline recommendations, exploring differential diagnoses). Independent survey data corroborates this directly: literature search is the single most common reported use case at 35% of physicians, followed by voice-based ambient documentation tools at 29%.
1.3 Which Tools Physicians Actually Trust
Physicians are not using one AI tool — they are using a portfolio, split between general-purpose assistants and healthcare-specific platforms. Reported usage: ChatGPT 83%, Google Gemini 50%, OpenEvidence 36%, and Wolters Kluwer's UpToDate with AI search 32%. This portfolio behavior matters strategically: a pharmaceutical brand's evidence has to be visible and correctly represented across a fragmented set of AI ecosystems, not a single dominant one, and some of those ecosystems (UpToDate) do not accept any form of paid or brand-originated content at all.
2. From Search to Ask: The New Information Ecosystem
For decades, scientific exchange followed a predictable, push-based model: pharmaceutical companies generated evidence, Medical Affairs interpreted it, Marketing distributed it through representatives, email, websites, and congresses, and healthcare professionals searched for it when they needed it. Success depended on reach and frequency. The challenge was distribution.
2.1 The Emerging Model: Intelligent Knowledge Access
Today, evidence increasingly reaches the healthcare professional through an AI system rather than directly from the pharmaceutical company: Evidence → AI System → Healthcare Professional; Guideline → AI System → Healthcare Professional; Medical Information → AI System → Healthcare Professional. Scientific exchange is becoming demand-driven — HCPs receive information when they need it, not when a company decides to deliver it. The challenge is no longer distribution. The challenge is discoverability and trust.
2.2 The Evolution of Scientific Exchange: Three Phases
| Phase | Focus |
|---|---|
| Phase 1 — Product Promotion | Reach, frequency, brand recall |
| Phase 2 — Scientific Engagement | Evidence, education, medical value |
| Phase 3 — Intelligent Scientific Exchange | Personalized knowledge delivery, AI-enabled evidence discovery, point-of-care support, real-time relevance |
Most organizations are still operating Phase 2 processes while their audience has already moved into a Phase 3 information environment. That gap — not a lack of scientific rigor or content volume — is the structural problem this whitepaper addresses.
3. What This Means for Pharmaceutical Companies
The role of the pharmaceutical company is expanding, not shrinking. Historically, commercial and medical success depended on clinical evidence, brand awareness, and commercial reach. In the AI era, success increasingly depends on evidence accessibility, content quality, knowledge structure, scientific credibility, and machine readability. The winners will not necessarily be those who produce the most content — they will be those who make their evidence easiest to understand, validate, and retrieve, for a human and a machine alike.
- Generative Engine Optimization (GEO): publish MLR-approved drug information in a structured, question-and-answer-oriented format aligned to physicians' actual clinical questions, so AI-powered search and answer engines can retrieve, interpret, and correctly cite it.
- Win in walled-garden medical ecosystems: audit which clinical AI platforms target physicians actually use, and ensure visibility across the underlying sources that feed them — medical journals, clinical trial registries, professional society publications — since platforms like UpToDate accept no brand-originated content at all, while others like OpenEvidence and Doximity accept targeted, disclosed pharma placements.
There is no single path to reaching physicians through AI. Consumer-facing assistants like ChatGPT and Gemini are exploring advertising models but are unlikely to permit direct prescription drug promotion soon; healthcare-specific platforms vary widely in what they will and will not accept. A brand's AI-visibility strategy has to be built platform by platform, not assumed to generalize from one success.
4. The AI-Ready Scientific Engagement Model: Five Pillars
Building an AI-ready scientific engagement model rests on five pillars. Each pillar corresponds directly to a capability travalcon has already built and documented as a standalone framework — the point of this section is to show that these are not five separate initiatives, but five interlocking parts of one response to one market shift.
| Pillar | What It Requires | travalcon Capability |
|---|---|---|
| 1. Evidence Excellence | Scientific content that is credible, current, source-linked, and structured — the foundation of trust | BCB Framework™ (Strategic & Analytics Layers) |
| 2. Content Intelligence | Modular knowledge assets — reusable, personalizable, updatable, consumable by both humans and AI systems | Modular Content Framework™ |
| 3. Medical Intelligence | Clinical studies, guidelines, real-world evidence, and medical information connected into one knowledge ecosystem | Knowledge Graph Framework™ |
| 4. AI Visibility | Understanding how AI systems represent disease states, treatment pathways, and evidence — and ensuring it stays discoverable and accurately represented | AI Visibility Optimization Framework™ |
| 5. Point-of-Care Value | Scientific support delivered where decisions actually happen — evidence navigators, medical copilots, guideline assistants | Personalization & Orchestration Framework™ + Tagging & Taxonomy Framework™ |
None of these pillars functions in isolation. Evidence Excellence without Content Intelligence produces trustworthy documents that cannot be reused or personalized. Medical Intelligence without AI Visibility produces a rich internal knowledge graph that no external AI system ever cites. AI Visibility without Evidence Excellence produces content optimized for citation that collapses under scientific or regulatory scrutiny. The five pillars are only as strong as their weakest, ungoverned link.
5. Five Strategic Opportunities for Pharma
Translated into initiatives a Medical Affairs or Commercial organization can actually fund and staff, the five pillars become five strategic opportunities.
| Opportunity | What It Delivers |
|---|---|
| 1. AI Visibility Optimization | Ensure scientific evidence can be found, understood, and referenced by AI systems; success increasingly depends on machine-readable evidence, not just human-readable evidence |
| 2. Medical Knowledge Platforms | Transform fragmented content into structured knowledge — connecting studies, guidelines, real-world evidence, claims, and patient populations into a unified evidence ecosystem |
| 3. AI-Powered Content Operations | Move from document creation to content orchestration — faster generation, better localization, improved reuse, lower production cost |
| 4. Point-of-Care Engagement | Provide value where decisions are made — evidence navigators, guideline explorers, clinical education assistants, medical information copilots |
| 5. Personalized Scientific Engagement | Deliver the right information to the right HCP, at the right time, through the right channel, in the preferred format |
Organizations that pursue these five opportunities together report lower content production costs, faster dissemination of evidence, improved engagement quality, better medical information support, greater scientific visibility, and stronger commercial performance. The competitive advantage belongs to the organizations whose evidence is easiest for both humans and AI systems to understand and use — not to the organizations that publish the most.
6. Case in Point: OpenEvidence and the Walled-Garden Medical Ecosystem
No single example illustrates the speed of this shift better than OpenEvidence, a clinical-AI platform frequently described as "the ChatGPT for doctors." Founded to give physicians a fast, citation-linked answer engine over the medical literature, the platform has moved from a niche tool to default clinical infrastructure in under two years.
| Metric | Figure |
|---|---|
| Verified physicians registered | 757,000+ |
| Daily active use among U.S. practicing physicians | 40%+, across 10,000+ hospitals and medical centers |
| Monthly clinical queries (January 2026) | 20 million+, up from ~18 million in December 2025 |
| Single-day consultation record | 1 million clinical consultations in 24 hours (March 10, 2026) |
| Valuation (January 2026 Series D) | $12 billion, on a $250 million raise led by Thrive Capital and DST |
| New-user referral source | 95% hear about the platform from another physician |
Two implications matter more than the growth figures themselves. First, OpenEvidence has announced a strategic collaboration with Veeva Systems to build an AI product suite connecting physicians and patients to clinical trials and improving adoption of approved therapies — a direct signal that clinical-AI platforms and pharma commercial infrastructure are converging, not staying separate. Second, OpenEvidence — unlike UpToDate — does accept targeted, disclosed pharma placements, making it one of the few large-scale clinical AI ecosystems where a brand can pursue both organic citation (through AI Visibility Optimization) and a direct placement strategy simultaneously. Treating every clinical AI platform as if it followed the same rules is already a measurable strategic error.
7. The Trust Imperative: Why Physicians Verify Before They Adopt
Despite high adoption, clinicians do not blindly trust AI outputs. Healthcare professionals consistently verify information against clinical guidelines, peer-reviewed literature, expert opinion, and institutional standards. The future is not AI replacing physicians. It is physicians working with AI — and that working relationship has a specific, demanding set of requirements.
- Accurate — clinically and scientifically correct, not just plausible
- Transparent — the reasoning and evidence base must be inspectable
- Evidence-based — traceable to primary sources, not synthesized opinion
- Source-linked — every claim carries a verifiable citation
- Easy to access — no friction between the question and the answer
- Integrated into workflow — available at the point of decision, not a separate destination
Healthcare is moving from search-driven information access toward AI-curated knowledge access. Tomorrow's clinician increasingly asks "What does the evidence suggest for this patient?" rather than "Where can I find the evidence?" Organizations that provide trustworthy, structured, evidence-rich information will matter more in this ecosystem, not less — but only if that content is built to survive the verification step every physician still performs.
8. The Medical Science Liaison in the AI Era
The Medical Science Liaison role is being reshaped by the same forces reshaping HCP information behavior, but not in the direction many feared. MSLs remain field-based Medical Affairs professionals who build trusted, non-promotional relationships with physicians and other healthcare experts to exchange scientific data and gather real-world insight. What is changing is how much of the surrounding work is now handled by AI before an MSL ever enters the room.
AI is increasingly used to automate literature scanning, analyze HCP engagement patterns, and manage scheduling and administrative overhead — freeing MSLs to spend more of their time in meaningful, science-driven conversations rather than preparation and paperwork. A newer category, sometimes called "AI MSLs" or agentic field-medical assistants, goes further: these systems interpret context, guide structured conversations, escalate appropriately, and generate structured insight reports, positioned explicitly to augment human MSLs rather than replace them. Leading MSL training curricula have already added dedicated 2026 modules on AI applications in Medical Affairs, covering research, insight analysis, content preparation, and strategic decision support — a clear signal that AI fluency is becoming a core MSL competency, not an optional add-on.
The implication for organizations building an intelligent scientific exchange model is direct: the same governed evidence base, knowledge graph, and tagging architecture that makes content citable to an external AI system is what makes an internal AI MSL assistant trustworthy to the humans using it. These are not two separate technology investments.
9. Governance: Why Scientific Engagement Cannot Be an Ungoverned AI Free-for-All
Every opportunity described so far carries the same risk if pursued without governance: an AI system citing an off-label claim, an outdated safety statement, or a superseded guideline is not a missed opportunity, it is a regulatory and patient-safety incident. Optimizing for AI citation without embedding MLR-grade governance at the point of content creation is not a shortcut — it is a liability generator at scale.
This is precisely the governance model travalcon's BCB Framework™, Tagging & Taxonomy Framework™, and AI Visibility Optimization Framework™ already embed by design: every claim carries an individual, verifiable source tag rather than a document-level disclaimer; every published asset carries machine-readable MLR metadata (author, medical reviewer, version, approval date); and human sign-off remains mandatory before publication regardless of how much of the drafting or citation-testing workflow is automated. Scientific engagement in the AI era does not require a new governance model bolted onto the old one. It requires the governance model regulated industries already know, applied to a new distribution channel that happens to be a language model rather than a sales representative.
10. Market Validation: The Numbers Behind the Shift
The scale of investment now flowing into this space corroborates the strategic argument on its own terms. The AI-in-medical-writing market is assessed at approximately $974 million in 2026, on a trajectory toward roughly $1.76 billion by 2030 at a 12.5% CAGR. The real-world evidence solutions market sits at $2.6–3.6 billion in 2026 depending on methodology, with AI-powered RWE solutions specifically forecast to grow at nearly 15% CAGR through 2035. OpenEvidence's jump to a $12 billion valuation in a single funding round is not an isolated data point — it is the clearest signal yet that investors believe clinical AI has become durable infrastructure, not a passing pandemic-era novelty.
- Physician AI adoption more than doubled in two years (38% → 81%+) — this is now majority behavior, not early-adopter behavior
- No physician specialty has adoption below 85% in the broader 7-country survey — this is not concentrated in a few AI-forward disciplines
- 73% of physicians say AI influences their clinical decisions to some degree — visibility inside AI systems is now a clinical-influence channel, not just an information channel
- Capital is flowing specifically into the platforms that sit between evidence and the physician (OpenEvidence, RWE tooling, AI medical writing) — the industry is investing in the layer this whitepaper is about
11. What Excellence Looks Like: Medical, Digital, and AI Excellence Combined
Leading organizations in this transition combine three forms of excellence simultaneously, not sequentially: Medical Excellence (high-quality, defensible evidence), Digital Excellence (omnichannel engagement infrastructure), and AI Excellence (structured knowledge and intelligent delivery). Any one alone produces a partial result — excellent science that never reaches the point of decision, or a sophisticated digital engine with nothing trustworthy to deliver.
Together, these three forms of excellence create a materially different HCP experience: faster time from published evidence to point-of-care availability, content that survives both a physician's manual verification and an AI system's retrieval and citation logic, and a measurable link between scientific engagement investment and prescribing or adoption behavior — the same behavioral-outcome discipline the BCB Framework™ already requires of every other content investment.
12. Illustrative Program Outcome
Consider a specialty biologics brand whose Medical Affairs and Commercial teams jointly commission an intelligent scientific exchange build-out ahead of a competitive label expansion. A Discover-equivalent audit finds the brand entirely absent from OpenEvidence and ChatGPT citations for its own therapeutic area, while a competitor's older, less differentiated safety data is being surfaced by default. A twelve-week Build phase restructures the brand's core evidence set into pyramid-chunked, source-tagged Knowledge Artifacts anchored to a newly built therapeutic-area knowledge graph, with MLR metadata embedded at the point of authoring rather than bolted on afterward.
By week twelve, re-testing against a Prompt Book of the therapeutic area's most common clinical questions shows the brand's own governed content now returned as the cited source on the majority of previously competitor-dominated queries, alongside measurable uptake of a companion point-of-care evidence navigator by field-facing MSLs. Continuing the cycle quarterly is what converts a one-time visibility gain into a durable advantage as the underlying evidence base and competitive landscape keep evolving.
13. Industry Applicability: Financial Services & Industrial B2B
While this whitepaper is written from a life sciences vantage point, the underlying shift — expert judgment increasingly mediated by an AI layer that must be trusted, verified, and correctly cited — is not unique to medicine.
| Industry | The Equivalent Shift | The Equivalent Response |
|---|---|---|
| Financial Services | Advisors and investors increasingly ask AI to compare products, risk, and suitability before a human conversation happens | Suitability-safe, per-claim-sourced answer content; the same five-pillar model applied to advisory and product evidence |
| Industrial / B2B | Engineers and procurement teams use AI to pre-qualify vendors and compare technical specifications before an RFP is issued | Technical evidence structured and tagged the same way clinical evidence is, so it is retrievable and citable by procurement-stage AI tools |
14. Competitive Benchmarking: Intelligence-Ready vs. Status-Quo Organizations
| Dimension | Status-Quo Organization | Intelligence-Ready Organization |
|---|---|---|
| Evidence structure | Long-form PDFs and slide decks, unstructured for machine retrieval | Pyramid-chunked, source-tagged Knowledge Artifacts, structured for both human and AI consumption |
| AI citation baseline | Absent or default to competitor/third-party sources | Actively measured Visibility Score, tracked per platform |
| MSL enablement | Manual literature review and prep; AI treated as a personal productivity tool at best | AI MSL / evidence navigator tools built on the same governed knowledge base as external-facing content |
| Governance model | MLR review applied inconsistently to AI-adjacent content, if at all | Per-claim source tagging and embedded MLR metadata applied uniformly across every channel including AI |
| Platform strategy | Single generic "digital" strategy assumed to cover all AI touchpoints | Platform-specific strategy (OpenEvidence vs. UpToDate vs. ChatGPT) reflecting each ecosystem's actual rules |
15. Organizational Readiness for the Transition to Intelligent Scientific Exchange
| Readiness Dimension | What Good Looks Like |
|---|---|
| Executive sponsorship | Joint Medical Affairs and Commercial ownership, not a Marketing-only or Medical-only initiative |
| Evidence governance maturity | MLR and medical review processes already capable of per-claim, source-tagged review rather than document-level sign-off |
| Knowledge infrastructure | An existing or planned knowledge graph and tagging taxonomy covering the therapeutic area's studies, guidelines, and RWE |
| AI platform visibility baseline | A completed audit of where the brand currently stands (or is absent) across the clinical AI platforms its target HCPs actually use |
| Field medical alignment | MSL leadership engaged early, so AI tooling is positioned as augmentation of the MSL role, not a threat to it |
16. Strategic Implications & Five Lessons from Early Movers
The central strategic implication is a reframed question for Medical Affairs and Commercial leadership. The question is no longer "How do we reach HCPs?" It is "Will our evidence be present, correctly represented, and trusted when an HCP asks AI for an answer?" That question cannot be answered by a single campaign, a single platform partnership, or a single content refresh. It requires the same operating discipline travalcon has already built for brand, content, tagging, knowledge, personalization, and AI visibility — applied together, for the first time, to the specific and urgent case of scientific exchange.
| # | Lesson |
|---|---|
| 1 | Physician AI adoption is majority behavior now, not a future scenario — the audit gap this reveals is usually larger than leadership expects |
| 2 | Trust is earned by structure and sourcing, not by content volume — physicians verify AI outputs against guidelines and literature regardless of adoption rate |
| 3 | No single AI platform strategy generalizes — OpenEvidence, UpToDate, ChatGPT, and Gemini each have different rules, different audiences, and different governance implications |
| 4 | Internal AI enablement (MSL tools) and external AI visibility (citation optimization) should share one governed evidence base, not be built as separate projects |
| 5 | The five-pillar model compounds — an organization that builds Evidence Excellence, Content Intelligence, Medical Intelligence, AI Visibility, and Point-of-Care Value together outperforms one that builds any single pillar in isolation |
Appendix: Intelligent Scientific Exchange Reference Architecture
- Before committing budget to an intelligent scientific exchange program, confirm: (1) Medical Affairs and Commercial share ownership and a common evidence base; (2) MLR/medical review can operate at the per-claim level, not just the document level; (3) a knowledge graph or structured content inventory exists (or is planned) for the therapeutic area in question; (4) a platform-by-platform AI visibility audit has been completed, not assumed; (5) MSL leadership is engaged before any AI field-tool is deployed.
Maturity Level Quick Reference
| Level | Characteristics |
|---|---|
| Level 0 — Unaware | No knowledge of the brand's current AI citation status; evidence unstructured; MSLs use consumer AI tools informally, ungoverned |
| Level 1 — Assessed | AI visibility audit complete across major clinical AI platforms; source-risk gaps identified; MLR governance model reviewed for AI readiness |
| Level 2 — Structured | Core evidence restructured into governed, tagged Knowledge Artifacts; therapeutic-area knowledge graph in build or in place |
| Level 3 — Visible | Measurable citation presence across priority AI platforms; MSL-facing evidence navigator or copilot tools in pilot |
| Level 4 — Compounding | Five-pillar model operating as one continuous cycle; quarterly re-measurement; MSL and external AI visibility sharing one evidence base |
Three-Phase Implementation Checklist
- Complete a platform-by-platform AI visibility audit (OpenEvidence, ChatGPT, Gemini, UpToDate, and any others relevant to the therapeutic area)
- Assess current MLR/medical review process for per-claim, source-tagged capability
- Inventory existing evidence: studies, guidelines, real-world evidence, medical information responses
- Secure joint Medical Affairs and Commercial executive sponsorship
- Engage MSL leadership on the intended role of AI tooling in field medical work
- Restructure priority evidence into pyramid-chunked, source-tagged Knowledge Artifacts
- Build or extend the therapeutic-area knowledge graph connecting studies, guidelines, RWE, and claims
- Apply machine-readable MLR metadata (author, reviewer, version, approval date) at the point of authoring
- Pilot an MSL-facing evidence navigator or medical information copilot on the same governed evidence base
- Determine platform-specific engagement rules (organic-only vs. disclosed placement) for each priority AI ecosystem
- Re-test AI citation presence against a defined Prompt Book on a recurring cadence
- Track leading indicators: citation share by platform, MLR cycle time, evidence reuse rate
- Track outcome indicators: MSL engagement quality, HCP engagement lift, prescribing or adoption signal
- Feed findings back into evidence prioritization for the next quarterly cycle
- Expand the model to additional therapeutic areas or brands once the first cycle demonstrates measurable impact
- 1. The question is no longer how to reach HCPs — it is whether your evidence is present, correct, and trusted when AI answers on your behalf.
- 2. Trust is structural, not promotional — physicians verify against guidelines and literature regardless of how convenient the AI answer is.
- 3. Internal AI enablement and external AI visibility are the same problem, solved once, from one governed evidence base — not two separate projects.
About This Whitepaper and travalcon.com
This whitepaper extends the strategic thinking captured in travalcon's internal concept for The Future of Scientific Engagement — the three-part narrative of how HCPs use AI today, what it means for pharmaceutical companies, and where scientific exchange is heading — with current 2026 market research on physician AI adoption, clinical AI platform growth, Medical Science Liaison evolution, and adjacent market sizing. It positions travalcon's six capability frameworks (BCB, Modular Content, Tagging & Taxonomy, Knowledge Graph, Personalization & Orchestration, and AI Visibility Optimization) as five interlocking pillars of one response to a single, well-documented market shift.
Figures cited from third-party research are attributed to their original sources; the underlying strategic framing — the shift from information push to intelligent knowledge access, and the five-pillar model — reflects travalcon's own methodology and internal concept work.