Evernorth announced a big bet on July 1: $100 million through 2028, aimed at a new AI program called Pharmacy Forward. The projected return is $400 million in value by the end of that window, starting with Accredo Specialty Pharmacy and its million-plus patients.
The mechanics are genuinely impressive. Prescription processing time cut in half. Clinician documentation time down by up to 50%. Predictive analytics flagging patients at risk of falling off therapy. AI choosing which of nearly 40 facilities ships a given medication, based on location, timing, and handling needs.
Here’s the question worth sitting with: which of those tasks belong to a machine, and which belong to a physician?
Routing a shipment from the nearest warehouse is a logistics problem. AI is built for exactly that. Pattern recognition at a scale that outpaces any human review, applied to a decision with low clinical stakes. RazorMetrics runs the same playbook. The RazorCodex scans millions of members, flags savings opportunities, and surfaces options at a speed only automation delivers.
Deciding whether a patient’s therapy needs to change is a different category of decision entirely. That call carries a person’s health on the other end of it, and it deserves a physician who knows the whole patient. The full history. The full context. Judgment that population-level patterns can approximate at best.
RazorMetrics draws that line deliberately. AI does the searching, the flagging, the scaling. Physicians make the call on what happens to an actual patient. That split is the reason a technology-forward pharmacy program stays a trustworthy one too.
Evernorth’s investment says something true about where the industry is headed: AI at scale is coming for every corner of specialty pharmacy, RazorMetrics included. AI’s place in the workflow is settled. What remains open is how much of that workflow AI should own. $100 million buys a lot of automation. The decision itself is worth keeping separate.
