Zachary Zubulake

Megha Sinha, Kolter AI, talks on how Orchestrator roles surged in pharma AI, but teams need computed plans, not titles, to close the skills gap.

Search list Midyear 2026 Check-In: Megha Sinha on Pharma's AI Skills Gap Midyear 2026 Check-In: Megha Sinha on Pharma's AI Skills Gap Jul 17, 2026, 12:00pm Midyear 2026 Check-In: Agentic AI Cuts Deviation Investigations by 50% Midyear 2026 Check-In: Agentic AI Cuts Deviation Investigations by 50% Jul 17, 2026, 12:00pm Midyear 2026 Check-In: AI Investment Skipped Pharma's Costliest Operational Problem Midyear 2026 Check-In: AI Investment Skipped Pharma's Costliest Operational Problem Jul 17, 2026, 12:00pm Brent Wilhelm on Cold Chain Pharmaceuticals Reshaping Shipping Needs Brent Wilhelm on Cold Chain Pharmaceuticals Reshaping Shipping Needs Jul 22, 2026, 7:00am Can AI Fix Pharma's Last-Mile Delivery Blind Spot? Can AI Fix Pharma's Last-Mile Delivery Blind Spot? Jul 20, 2026, 7:00am Brent Wilhelm on Why Drug Shortages Persist Despite Available Inventory Brent Wilhelm on Why Drug Shortages Persist Despite Available Inventory Jul 16, 2026, 1:35pm From Predictions to Reality: Mike Stenberg Assesses Pharma Manufacturing's First Half of 2026 From Predictions to Reality: Mike Stenberg Assesses Pharma Manufacturing's First Half of 2026 Jul 17, 2026, 1:01pm Can Governance-by-Design Make AI Compliant in Pharma? (Part II) Can Governance-by-Design Make AI Compliant in Pharma? (Part II) Jul 16, 2026, 11:14am Overcoming Raw Material Variability Challenges for Autologous Adoptive Cell Therapies  Overcoming Raw Material Variability Challenges for Autologous Adoptive Cell Therapies Jul 08, 2026, 10:00am FDA Activity Round Up – July 2026 FDA Activity Round Up – July 2026 Jul 15, 2026, 9:00am The Implications of the FDA’s New Finalized Guidance Documents The Implications of the FDA’s New Finalized Guidance Documents Jul 16, 2026, 9:04am Midyear 2026 Check-In: AI, Supply Chains, and the NDC-12 Story Nobody's Watching Midyear 2026 Check-In: AI, Supply Chains, and the NDC-12 Story Nobody's Watching Jul 16, 2026, 7:44am Manufacturing Intelligence: Sizing Up Pharma's Biggest AI Deals, Part One Manufacturing Intelligence: Sizing Up Pharma's Biggest AI Deals, Part One Jul 15, 2026, 9:17am Using Risk-Based Quality Frameworks to Fast-Track Biopharma Development: A Q&A with Abzena’s Jeffrey C. Mocny and Cellares’ Anna McMahon Using Risk-Based Quality Frameworks to Fast-Track Biopharma Development: A Q&A with Abzena’s Jeffrey C. Mocny and Cellares’ Anna McMahon Jul 15, 2026, 9:25am Engineering High-Speed Pen Injector Final Assembly Machine: Solving the Cartridge Press-Fit Challenge Engineering High-Speed Pen Injector Final Assembly Machine: Solving the Cartridge Press-Fit Challenge Jul 14, 2026, 4:56pm EMA and EU Innovation Agency Deepen Ties to Speed Up Drug Development EMA and EU Innovation Agency Deepen Ties to Speed Up Drug Development Jul 15, 2026, 8:16am Mid-Year Check-In: Are Henrik Johanning's Pharma Manufacturing Predictions Holding Up? Mid-Year Check-In: Are Henrik Johanning's Pharma Manufacturing Predictions Holding Up? Jul 15, 2026, 8:12am It’s Time to Build a More Scalable, Connected CRO It’s Time to Build a More Scalable, Connected CRO Jul 09, 2026, 8:27am Hidden Cost Drivers in Oligonucleotide Manufacturing Hidden Cost Drivers in Oligonucleotide Manufacturing Jul 14, 2026, 10:55am Regulatory Actions Impacting Biologics Regulatory Actions Impacting Biologics Jul 09, 2026, 3:57pm Orally Dissolving Films: A Modern Solution for Convenient Drug Delivery Orally Dissolving Films: A Modern Solution for Convenient Drug Delivery Jun 24, 2026, 2:05pm What FDA's AI Warning Letter Tells Us About GMP Accountability What FDA's AI Warning Letter Tells Us About GMP Accountability Jul 06, 2026, 9:01am Munnik Revisits His 2026 Predictions: A Midyear Check-In on Regulatory Reform, Data, and AI in Pharma Munnik Revisits His 2026 Predictions: A Midyear Check-In on Regulatory Reform, Data, and AI in Pharma Jul 13, 2026, 12:01pm Using Automation to Enhance Pharmaceutical Manufacturing Using Automation to Enhance Pharmaceutical Manufacturing Jul 13, 2026, 12:49pm Midyear 2026 Check-In: Megha Sinha on Pharma's AI Skills Gap

Access Part 1 of this interview on how digital twins run pharma supply chains but miss the regulatory execution needed to launch changes across market.

In part 2 of a 3-part interview, Megha Sinha, founder and CEO of Kolter AI

, returns from earlier in the year to speak with 

 about how the industry has developed in the first half of 2026. She turns to the skills gap behind pharma's AI rollout. The orchestrator role, she says, has gained the most visible traction across the industry, but too often it's been title-deep rather than skill-deep. "People got called the orchestrators without being taught how to orchestrate or use the engine to help them," Sinha says. That gap has been especially stubborn in lifecycle change execution and geographic expansion, where no system existed to generate a plan for teams to challenge. "You cannot train a team to pressure test a plan for site transfer across 40 markets when no system generates that plan," she notes, arguing the absence of a tool, not a lack of training, hid the real problem.

Sinha says that is now shifting as pharma companies bring engines like Kolter AI into live programs, computing an entire cross-market change, including filings, label updates, and country-by-country implementation waves, before a single affiliate is contacted. Local teams then validate a computed plan rather than building one from scratch over months, a shift she frames as a change in the job itself, from constructing plans to interrogating them.

She also pushes back on industry timelines that place connected, cross-functional orchestration a decade out. "The connective tissue that everyone described as years away is exactly what we built," Sinha says, describing the missing piece as automated impact assessment spanning regulatory, quality, manufacturing, and labeling functions. Companies approaching Kolter AI, she says, aren't asking whether this kind of orchestration is possible after years of doing it manually. They're asking how quickly they can adopt it, which Sinha takes as confirmation that the shift from aspiration to operational reality is already underway.

Videos

Megha Sinha, Kolter AI, talks on how orchestrator roles surged in pharma AI, but teams need computed plans, not titles, to close the skills gap.

Midyear 2026 Check-In: Megha Sinha on Pharma's AI Skills Gap

Bristol Myers Squibb announced plans to expand its computing infrastructure collaboration with NVIDIA

Bristol Myers Squibb (BMS) has announced plans on July 20, 2026 to expand its computing infrastructure with NVIDIA DGX Vera Rubin NVL72 systems, a move the company says will give it the most advanced and energy-efficient artificial intelligence (AI) infrastructure of any single owner in life sciences.

The deployment extends a collaboration that began nearly 3 years ago, when BMS first adopted NVIDIA DGX SuperPOD systems to support its research and development work.

The Vera Rubin architecture is designed to deliver up to 10 times greater performance per megawatt than its predecessor.

 That efficiency gain matters as much as the raw processing boost, because it allows BMS to pursue more complex AI models without a proportional rise in energy demand. The expanded infrastructure will support research programs across oncology, hematology, cardiovascular disease, immunology, and neuroscience.

Why Does Computing Capacity Matter to Drug Developers?

Pharmaceutical development increasingly depends on the ability to train and run proprietary models on large volumes of internal experimental and clinical data.

 The scale of that computing capacity shapes how quickly companies can test hypotheses, design molecules, and interpret results, which in turn affects how efficiently a program moves from discovery toward clinical development. As more of the industry builds this kind of dedicated capacity, computational infrastructure is becoming a factor in development timelines alongside laboratory and clinical execution.

BMS points to 2 examples of this shift already underway.

 AI agents used for target identification and validation are reducing manual work for scientists. Separately, the company's Predict First approach uses model-generated predictions to inform experimental design before laboratory work begins, a method it says now informs every small molecule program and most large molecule programs it runs.

What Role Is AI Expected to Play Alongside Researchers?

BMS describes its broader goal as hybrid intelligence, an approach in which computational systems handle data-intensive execution while human researchers retain responsibility for direction, interpretation, and judgment calls that require deeper expertise.

 The Vera Rubin cluster is intended to serve as the computational foundation for that model, supporting foundation models trained on the company's proprietary research data and, where applicable, incorporating domain-specific tools from NVIDIA's biological AI platform, BioNeMo.

“Drug discovery is a sequence of decisions made under uncertainty, and better decisions come from better evidence, faster,” said Robert Plenge, executive vice president and chief research officer, Bristol Myers Squibb, in a press release.

 “This infrastructure lets us learn from every experiment and every clinical readout to sharpen the next hypothesis, allowing BMS scientists to spend less time on manual work and more time on the questions that require human judgment. The goal isn't speed for its own sake; it's raising the probability that each program we advance is the right one.”

The significance lies less in the hardware itself and more in what it signals about where competitive advantage is shifting.

 As computational tools become embedded earlier in discovery and design, organizations that can evaluate more hypotheses, faster and with higher confidence, may be positioned to bring a narrower, better-vetted set of candidates into costly development and manufacturing pipelines. That could influence how resources are allocated downstream, from process development through scale-up, as fewer but more thoroughly validated programs move forward.

How Big Could This Infrastructure Shift Become?

 Analysts at Global Reports Store value the global AI drug discovery infrastructure market at $3.10 billion in 2025, projecting growth to $19.98 billion by 2032. That trajectory reflects a broader shift: pharmaceutical research is being rebuilt around compute, data, and generative molecular design rather than isolated software tools. Vendors and internal groups are increasingly judged on whether their data architecture, computing environment, and lab integration can make artificial intelligence useful at production scale, not just in early discovery.

Bristol Myers Squibb to build the most powerful AI factory in life sciences with NVIDIA. Press release. Bristol Myers Squibb. July 20, 2026. 

https://news.bms.com/news/corporate-financial/2026/Bristol-Myers-Squibb-to-Build-the-Most-Powerful-AI-Factory-in-Life-Sciences-with-NVIDIA/default.aspx

Global Reports Store. AI Drug Discovery Infrastructure Market to Add US$16.88 Billion by 2032 as Pharma, Cloud and Bio-AI Platforms Rebuild R&D Around Compute, Data and Generative Molecular Design. 

https://www.openpr.com/news/4512781/ai-drug-discovery-infrastructure-market-to-add-us-16-88-billion

Articles

Bristol Myers Squibb plans to expand its in-house AI compute with NVIDIA DGX Vera Rubin NVL72 systems, targeting a step-change in performance per megawatt and positioning compute efficiency as a core R&D lever. The infrastructure is intended to scale foundation models trained on proprietary experimental and clinical datasets, supporting hybrid intelligence workflows where machines execute data-heavy tasks and scientists guide interpretation. Use cases include AI agents for target ID/validation and a “Predict First” paradigm informing most discovery programs. The move reflects an industry-wide shift toward compute-centric drug development.

BMS is scaling AI drug discovery with new NVIDIA infrastructure, cutting research timelines via faster, more efficient computing power.

Everything to Know about BMS & NVidia’s AI Supercomputer

Brent Wilhelm, Cencora talks on last-mile drug delivery risks, carrier network visibility, weather disruptions, and how AI could improve routing.

Can AI Fix Pharma's Last-Mile Delivery Blind Spot?

Access part 1 of this interview on why pharma supply chains still run on historical demand data and how AI could close the real-time visibility gap.

In part 2 of a 3-part interview, Brent Wilhelm, Senior Vice President of Global Supply Chain, Cencora, turns to the final mile: getting pharmaceuticals from the distribution network into the hands of practitioners and patients. Even when a therapy is in stock and positioned correctly, Wilhelm notes, delivery in the US typically depends on third-party carriers like FedEx and UPS, which introduces risk beyond Cencora's direct control. "And so where there's a larger risk is having visibility into the health of those networks," he says, pointing to the need to understand whether carriers can reliably pick up and deliver shipments on schedule.

That visibility gap becomes acute during disruptions. Wilhelm cites this past winter's Winter Storm Fern as a case in point: "Winter Storm Fern this past winter was a big issue in the southeast," he says, describing how the storm affected FedEx's Memphis hub and UPS's Louisville hub for days. Shipping from the geographically closest distribution site doesn't always make sense, he explains, if that carrier lane is experiencing latency or weather-related delays, which is why Cencora is exploring AI to consume real-time carrier data and route shipments through the healthiest available network path.

Wilhelm also highlights how delivery cadence varies by site of care. Some treatment centers need shipments once daily, others multiple times a day, and turnaround expectations range from next-day to two-day fulfillment depending on the therapy and practitioner. Managing that complexity, he says, requires both a flexible distribution footprint and the operational capability to support varied delivery methods simultaneously. Cencora's network of distribution sites is built to flex to those differing customer needs.

Together, the insights underscore a theme from Wilhelm's broader conversation: that resilient pharmaceutical delivery increasingly depends on combining real-time data with AI-driven decision-making, rather than relying solely on historical shipping patterns or proximity-based routing.

Brent Wilhelm, Cencora, talks on last-mile drug delivery risks, carrier network visibility, weather disruptions, and how AI could improve routing.

Megha Sinha, Kolter AI, discusses how digital twins run pharma supply chains but says they miss the regulatory execution needed to launch changes across markets.

 Midyear 2026 Check-In: Agentic AI Cuts Deviation Investigations by 50%

Megha Sinha, founder and CEO of Kolter AI and Kamet Consulting Group, 

spoke with PharmTech at the beginning of 2026 about her predictions for the industry this year.

 Now that the year is halfway through, In part 1 of a 3-part interview, she discusses how she believes that the infrastructure around digital twins is "Half built, and the built half is the easier half."

Companies such as Merck, Novartis, and BMS now run real-time what-if simulations on select S&OP processes, and the market for plant-level digital twins has surpassed a billion dollars. Inside the plant, agentic systems are already detecting out-of-spec results, pulling batch records, drafting root cause analyses, and cutting investigation time by 50% or more in validated GxP environments. "That is a shift from AI as a productivity layer to AI as an operational participant," Sinha notes.

But Sinha argues these systems stop short of where the real cost lies: translating a network, process, or manufacturing change into its regulatory consequences market by market. A minor change on the plant floor, she explains, can trigger a major variation requiring prior approval in Brazil, a six-month grace period in Germany, or a bundling opportunity across other markets, none of which existing digital twins or agentic tools are built to compute. Tariff planning, she adds, is still largely a quarterly review exercise rather than a live decision engine.

That gap between physical execution and regulatory execution is what Sinha says Kolter AI was built to close. "To me, that is the next frontier, and companies should be moving on it now," she says, describing an engine that takes a network or process decision.

Megha Sinha, Kolter AI, discusses how digital twins run pharma supply chains but  miss the regulatory execution needed to launch changes across markets.

Midyear 2026 Check-In: Agentic AI Cuts Deviation Investigations by 50%

Brent Wilhelm, Cencora discusses why pharma supply chains still run on historical demand data and how AI could close the real-time visibility gap.

Brent Wilhelm on Why Drug Shortages Persist Despite Available Inventory

In part 1 of a 3-part interview, Brent Wilhelm, Senior Vice President of Global Supply Chain, Cencora, brings more than 30 years of healthcare industry experience to a role focused on getting pharmaceuticals from manufacturers to patients quickly, including specialty and temperature-controlled products, across both domestic and global networks.

Wilhelm explains that Cencora's US network relies on more than 25 distribution centers designed to place inventory close to customers based on anticipated demand, enabling delivery in a matter of hours. But that model runs into trouble when demand shifts unexpectedly, as it did during regional RSV outbreaks. "So when we have demand variability, moving inventory, although it's available somewhere in the network, it could be on the East Coast, but the demand's spiking on the West Coast," he says. Because repositioning inventory takes time, drugs can technically be in the supply chain yet unavailable where they're needed most.

The fix, Wilhelm argues, is closer integration between distributors and manufacturers so that real patient-level demand and dispensing data can inform planning decisions before shortages occur. That shift is increasingly urgent as treatments move toward more real-time administration and practitioners look to stock only what they need immediately rather than carrying inventory for future weeks. "The way that the supply chain is set up today across the US, it's much more of a historical demand, not a real time," he notes, pointing to a gap between how the system was built and how it now needs to function.

Wilhelm also sees artificial intelligence as essential to closing that gap, both for spotting demand signals in ordering patterns and for managing complexity around dosage strengths and formulations. A persistent blind spot, he says, is visibility into what's being dispensed at the point of care, a limitation Cencora is now discussing with larger customers as it works toward shared demand planning that protects supply chain resilience and patient safety.

Brent Wilhelm, Cencora, discusses why pharma supply chains still run on historical demand data and how AI could close the real-time visibility gap.

Topical Corticosteroid Bioequivalence Guidance Ends 30-Year Wait

The FDA finalized two guidance documents on July 14, closing out review periods that in one case stretched back more than three decades.

 Together, the documents address how sponsors interact with the agency during meetings for over-the-counter monograph drugs, and how generic drug developers should demonstrate that topical corticosteroid products perform the same way as their reference products. Neither document changes what data developers must ultimately generate, but both settle open questions that companies have had to navigate around for years, sometimes relying on outdated methodology or informal agency feedback in the absence of a current final policy.

What Does the New Bioequivalence Guidance Change for Generic Developers?

The corticosteroid guidance, Topical Dermatologic Corticosteroids: In Vivo Bioequivalence, replaces a document the agency first issued in 1995 and updates methodology for pharmacodynamic studies used to support abbreviated new drug applications (ANDAs).

It applies across all potency groups of topical corticosteroids and lays out the agency's current thinking on structuring a pilot dose-duration vasoconstrictor response study followed by a pivotal vasoconstrictor bioequivalence study, the pharmacodynamic approach long used as a surrogate for demonstrating that a generic topical corticosteroid delivers the same clinical effect as the branded product.

Compared with the draft version issued in 2022, the final guidance clarifies that subjects used to qualify a chromameter and its operator do not need to be the same subjects enrolled in the pilot or pivotal studies, and it adds a history of hypopigmentation to the list of subject exclusion criteria.

 For formulation and clinical teams designing an ANDA bioequivalence program, these are the kinds of details that determine protocol structure and subject screening well before a study begins. A clear, current reference reduces the chance of agency feedback arriving only after a study is already underway.

Why Does a Meeting-Procedures Guidance Matter to Product Developers?

The second document, Formal Meetings Between FDA and Sponsors or Requestors of Over-the-Counter Monograph Drugs, finalizes a 2022 draft and describes how sponsors request and conduct meetings tied to monograph order submissions under a pathway added to federal drug law in 2020.

 It sets out three categories of what the guidance calls Type Y meetings, generally limiting sponsors to one meeting per category over the course of a monograph order development program, and it accommodates in-person, virtual, teleconference, and written-response formats.

The final version also extends the guidance to cover meetings tied to clinical investigations conducted under an investigational new drug application, where the goal is to support a determination that a nonprescription drug is generally recognized as safe and effective, closing a gap sponsors had raised during the comment period.

 It also replaces an earlier term, OTC monograph industry working group, with joint meeting requester, to reduce confusion when multiple sponsors coordinate a single submission.

For companies managing over-the-counter monograph portfolios, meeting access and timing directly affect how quickly a monograph order can move forward, and the guidance ties these procedures to updated performance goals under the current monograph drug user-fee program, covering fiscal years 2026 through 2030.

 Predictable meeting pathways make it easier to plan submission timelines and allocate regulatory affairs resources across a development program, rather than negotiating meeting scope case by case.

Both guidances are now in effect as the agency's current thinking on these topics, though the FDA notes that sponsors may use an alternative approach if it satisfies the underlying statutory and regulatory requirements.

Topical Dermatologic Corticosteroids: In Vivo Bioequivalence; Guidance for Industry; Availability. Federal Register. 2026;91(133):43100-43101. July 14, 2026.

Formal Meetings Between FDA and Sponsors or Requestors of Over-the-Counter Monograph Drugs; Guidance for Industry; Availability. Federal Register. 2026;91(133):43098-43100. July 14, 2026.

FDA finalized two long-awaited guidances clarifying expectations for generic topical corticosteroid bioequivalence and formal meeting procedures for OTC monograph drug programs. The corticosteroid document modernizes pharmacodynamic vasoconstrictor study design for ANDAs, detailing pilot and pivotal approaches and refining operational elements that affect protocol build and screening. The OTC monograph meetings guidance standardizes how sponsors request and conduct Type Y meetings across submission stages, expands scope to certain IND-supported investigations, and aligns meeting access with OMUFA performance goals for FY2026–2030.

FDA finalizes bioequivalence guidance for topical corticosteroids and new meeting rules for OTC monograph drug sponsors, replacing a 1995 standard.