Author: Tim Manuel, Staff Product Manager at Claritas Rx
The Claritas Rx product team spent three days at this year’s annual specialty pharmacy conference, walking the exhibit floor, attending sessions across the AI, patient access, and tech-enabled care tracks, and comparing notes with manufacturers throughout the week. This one stays close to the product floor: what we heard on stage, on the floor, and in conversations throughout the week, what’s changing, and what it means for the specialty analytics roadmap.
Four themes stood out.
1. The AI Question Changed: From “Should We Invest” to “How Do We Invest”
A year ago, the recurring question on stage was whether manufacturers were investing in AI yet. This year, that question is settled. Survey data presented at the conference on the AI and real-world data capabilities pharma now considers essential to compete on access revealed two clear patterns. First, the top-ranked capabilities are operational; manufacturers want AI embedded in workflows and in the evidence flywheel. Second, AI-enabled patient journey orchestration ranked near the top, placing the orchestration layer squarely in the AI conversation rather than adjacent to it.
The governance counter-narrative: The investment story is only half the picture. The other half, heard repeatedly in manufacturer conversations, is governance. Several manufacturers have hard scope restrictions on AI/ML pending case-by-case approval. Others required AI capabilities to be explicitly scoped to specific patient populations, with certain payer categories off-limits. The takeaway isn’t that AI demand has cooled; it’s that “AI-powered” now needs to come paired with scoping controls, audit trails, and a governance posture that lets compliance teams say yes.
Therapy class shapes appetite: AI receptivity also varied sharply by therapy type. Cell and gene therapy manufacturers, where each enrollment is a high-stakes, high-cost, one-time decision, were the most bullish. Chronic and acute therapy manufacturers were materially more cautious, and some pushed back on the predictive paradigm itself: for episodic acute medications, “dropout” isn’t well-defined, which breaks the standard adherence-prediction frame. The implication is clear: AI value propositions need to be segmented by therapy class, not generalized across the book of business.
What it means for market access: The bar for what “AI-powered” means in this category is rising fast. Vendors who lead with use cases like predictive access analytics, near-real-time triggers, and journey orchestration will have more credibility than those leading with feature checklists. And capability alone won’t close deals: governance is increasingly table stakes for enterprise buyers, not a nice-to-have.
2. Patient Support Is Flipping From Outside-In to Inside-Out
The most product-resonant session of the conference made a compelling case: three decades of hub-centric patient support has hit a ceiling. Modeling an “inside-out” approach, pushing services into the EHR, e-Rx, switch, and pharmacy workflows where the patient journey actually happens, projects meaningfully higher dispense rates, with a significant portion of currently lost scripts and PA-stuck scripts recoverable.
The prescription was concrete: invest in studying the pipes. The pipes, including clinician inputs, e-prescribing, EHR, hub, pharmacy, and switch, are where the data trail lives and where actionable triggers can be placed. The future model is a real-time signal delivered to the right person at the moment of friction, not a quarterly flat file.
The institutional specialty pharmacy visibility gap: A concrete example of broken pipes surfaced repeatedly in manufacturer conversations: institutional specialty pharmacies (ISPs), including those embedded inside large hospital systems, frequently don’t share patient data with manufacturers. Brand teams have no visibility into PA status, appeals, or disposition for patients flowing through these channels, even on therapies with significant ISP utilization. The inside-out thesis is colliding with reality: the pipes exist, but not all of them are wired up.
What it means for market access: The ISP visibility gap is not a niche problem; it came up repeatedly and across different manufacturer conversations, which signals it is reaching a tipping point. The operative question for any analytics or data provider is which partners can actually get data out of the EHRs and workflow systems where the patient journey lives. Connectivity is becoming the differentiator.
3. Specialty Pipeline Pressure Is Rewriting the Access Equation
Even when “specialty pipeline” wasn’t the named topic, it shaped every session. Drug lifecycles are compressing. Biologic, cell, and gene therapies continue producing assets that need coverage. Biosimilar entry is accelerating. And payer aggressiveness, including denial rates, step therapy, and formulary exclusions, is rising in parallel.
Survey data presented at the conference showed that only one in six manufacturers still sees price as the primary access lever. The other 84% are working utilization management, contracting, value demonstration, and provider economics, meaning the access conversation is being pushed upstream into pre-launch and lifecycle planning, not held at launch and re-priced later.
What it means for market access: Market access analytics needs to keep up with that shift. Channel and gross-to-net reporting is necessary but no longer sufficient. The story manufacturers want to tell, covering UM positioning, predicted approval and abandonment risk, value-based contracting support, and real-world access outcomes, has to map cleanly onto the analytics surface, with account- and HCO-level rollups as a first-class view. The integrated picture across formulary, channel, GTN, and contracting is still open territory.
4. Agentic AI Is Eating the Specialty Pharmacy Back Office
Across the conference floor, the single hottest sub-theme was purpose-built agentic AI for specialty workflows: agentic operating systems for intake, benefits verification, prior auth, claims, and appeals; healthcare vision-language models; reinforcement-learning agents; multi-agent systems for prior auth; and voice agents calling payers autonomously. The category is moving fast; most of these companies have emerged just in the last two years.
Two flavors of agentic worth distinguishing: Most floor pitches are reasoning-and-orchestration agents that read documents, run rules, and coordinate steps across a workflow. A second flavor surfaced in off-floor conversations: operational agents that drive third-party UIs as if they were human users, sidestepping the cost and timeline of formal API integrations entirely. For the long tail of small SPs, payer portals, and ISP systems where integration economics don’t pencil out, this is a credible path to extending automation past where standard integration would stop. The trade-offs are real, including brittleness, exception handling, and third-party consent, but for accounts gated on integration cost, it’s a tool worth having.
What it means for market access: The back-office automation layer between manufacturers and their data is being rebuilt. For brands, the implication is clear: analytics need to fit on top of these workflow automation layers, not sit beside them. And as agentic AI moves into patient-facing workflows, the bar for what good voice and document handling looks like is rising across the board.
Where This Lands
The four themes converge on one product imperative: the analytics platform that wins the next cycle is the one closest to the workflow. Aggregated channel data and post-hoc dashboards are table stakes. The differentiation is real-time signal, embedded triggers, account and HCO-level pattern detection, and the integration footprint across pipes, EHR, payer systems, voice channels, and the long tail where standard integration breaks down, that makes the signal actionable. It also has to ship with credible AI governance: scoping controls, case-by-case enablement, and a posture that gives enterprise compliance teams a clear path to yes.
If any of these themes intersect with questions you’re sitting on for your brand, we’d love to compare notes and discuss further. Schedule A Meeting
Authored by the Claritas Rx Product Team based on AXS26 sessions, exhibitor conversations, and manufacturer interviews. Data points are speaker-cited and directional. Manufacturer commentary is intentionally presented without attribution.
