David Pagan

Even within a single company, sites differ in processes, equipment, systems, risks and regulatory requirements, so AI deployments have to account for those differences at the site level.

• They warn that when companies ignore that context, the consequences can reach the supply chain: shortages of goods or supplies, or worse, of drugs and therapies.
• They see the slow pace of AI adoption in manufacturing as potentially a good sign: the complexity demands care before scale.

From the AIMST recap: "Manufacturing isn't one homogeneous environment.

• Even within the same company, sites have different processes, equipment, systems, risks, and regulatory requirements.
• AI deployments need to take all these things into account."

Read original

In their view, business-critical software cannot depend on the executive who built it, one IT person who maintains it or undocumented grassroots knowledge.

• They draw on a familiar manufacturing idea: plants already understand single points of failure in physical operations.
• The question is how to apply that thinking to the software running alongside them.
• Software reliability, they suggest, becomes manufacturing reliability.

The post lists the dependencies it considers disqualifying — the executive builder, the lone IT maintainer, undocumented grassroots knowledge — and adds, "I won't even mention SDLC."

Read original

The next step, they argue, is knowledge-driven decision-making: combining operational data with the SOPs, policies, procedures and other information that define how manufacturing is supposed to work.

• They acknowledge this is partly a question of wording, but say the concept holds on its own.
• They also urge sites not to wait until all their data is perfectly structured before starting AI initiatives, and argue that AI in manufacturing operations is not a strategy by itself.

Pagan names the systems being tied together — ERP, MES, QMS, PLC and SCADA — and contrasts data-driven decisions with knowledge-driven decisions based on the procedures that govern production.

• The post closes with the line "For manufacturing operations: AI is not a strategy.
• That being said, what are you doing with AI?"

Read original

Across drug discovery, materials science and battery development, they saw a consistent opportunity: better modelling and understanding of molecular interactions that are difficult, expensive or time-consuming to explore through traditional approaches.

• AI already accelerates parts of that work, and quantum computing is increasingly discussed as an additional layer.
• But Pagan says most of the conversation focused on scientific and technological breakthroughs, while the path from discovery to commercial reality still runs through manufacturing operations.
• Speed to market, financial viability, regulatory compliance and scalability determine whether innovation creates value or remains an interesting result.

Pagan singles out a venture-capital comment made in the room — "The best technology does not always win" — as a point about execution, including go-to-market.

• They also record the mix of participants: IBM, Silicon Quantum Computing, Oxford, UCL, Moderna, Cleveland Clinic and Sanofi, among others.
• Pagan asks whether it is too early to invest in quantum, answers "maybe," and notes that countries, ecosystems and companies are already betting on it.

Read original

They describe Murmuratto, the company they work with, as a multinational with three offices, including one in San Juan, Puerto Rico.

• Its team includes bio experts with decades of regulated manufacturing experience at global biopharma and biotech sites — people who understand quality systems, regulatory requirements and production pressures.
• Pagan says that expertise shapes governance, deployment strategy and system architecture: the team designs around the jobs to be done, the workplace environment and how decisions actually get made.
• The stated design priority is the lowest barrier to entry and highest usability for people on the plant floor.

At BIO 2026 in San Diego, Pagan joined the Invest Puerto Rico delegation alongside Murmuratto's San Juan presence and wrote: "We're not designing great tools in a vacuum." They tie the Puerto Rico team directly to the industry, regulatory and practical manufacturing experience that shapes how the company builds.

Read original

They attend an industry gathering, identify the theme that kept resurfacing, develop it through numbered observations or a point of view, and end with a question for readers.

• The events range from manufacturing AI to quantum commercialisation to regional tech and biopharma gatherings.
• This gives readers a view of how Pagan forms and states positions publicly, anchored to something specific said or shown in the room rather than general commentary.

The AIMST recap is organized as three numbered takeaways and closes with a question to the audience.

• The Commercialising Quantum post identifies "use cases" as the theme that kept resurfacing and ends by asking whether it is too early to invest in quantum.
• A short ConvergeSouth note thanks the organizers and flags a State of the South keynote, which Pagan identifies as a guest's talk.

Read original · Original 2 · Original 3 · Original 4

It answers “what does this person actually think about deploying AI in manufacturing?” with three concrete claims, not general enthusiasm, making it the best starting point.

Pagan attended AI in Manufacturing and came away with three points.

• First, manufacturing is not one environment: even within a single company, sites differ in processes, equipment, systems, risks and regulatory requirements, so AI deployments must account for that variation.
• Pagan argues that ignoring this context risks shortages of goods, supplies, drugs or therapies.
• Second, software reliability can become a manufacturing reliability risk.
• AI makes it easy to prototype internal applications, but, Pagan says, business-critical software cannot depend on the executive who built it, the one IT person who maintains it or undocumented grassroots knowledge.
• Manufacturing already understands single points of failure; Pagan asks how that lesson applies to technology.
• Third, manufacturing should move from data-driven to knowledge-driven decisions.
• Tying ERP, MES, QMS, PLC and SCADA data together supports data-driven decisions; knowledge-driven decisions combine that data with the SOPs, policies and procedures that define how manufacturing is supposed to work.
• Sites should not wait for all their data to be perfectly structured before starting.
• Pagan closes with “AI is not a strategy.”

For brands planning a campaign

This post gives a talk or advisory session a ready structure: site variability, software reliability and knowledge-driven operations are three agenda items a regulated-manufacturing audience can act on. For a life-sciences technology company, it also tests fit: Pagan’s framing speaks to operations, quality and regulatory audiences, not just data-science teams.

Read this for the perspective Pagan would bring to a quantum or AI conversation in the life sciences.

Pagan reports from Commercialising Quantum, hosted by The Economist, with IBM, Silicon Quantum Computing, Oxford, UCL, Moderna, Cleveland Clinic, Sanofi and others in the room.

• The theme that kept recurring was "use cases."
• Across drug discovery, materials science and battery development, Pagan describes a consistent opportunity: improving the ability to model and understand molecular interactions that are difficult, expensive or time-consuming to explore by traditional means.
• AI already accelerates parts of that work, and quantum computing is increasingly discussed as another layer that could expand what is possible.
• Pagan's own emphasis is that much of the conversation focused on scientific and technological breakthroughs, while getting from discovery to commercial reality still depends on manufacturing operations.
• Speed to market, financial viability, regulatory compliance and scalability are the constraints that determine whether innovation creates value or stays an interesting result.
• Pagan singles out a venture-capital comment, "The best technology does not always win," as a reference to the importance of execution, including go-to-market, and asks which teams will win.
• Pagan asks whether it is too early to invest in quantum, says maybe, but notes that countries, ecosystems and companies are already betting on it.

For brands planning a campaign

For a life-sciences technology company assessing Pagan as a speaker, this supplies a clear thesis: the gap between breakthrough and execution matters as much as the breakthrough. That thesis can anchor a talk or advisory discussion for an audience weighing quantum and AI investments in regulated environments.

Read it to understand their working context and how they connect domain expertise to product decisions.

Pagan joined Invest Puerto Rico at BIO 2026 in San Diego and describes Murmuratto as a multinational company with three offices, including one in San Juan, Puerto Rico.

• The team includes bio experts with decades of regulated manufacturing experience at global biopharma and biotech sites, Pagan says.
• They understand the implications, challenges and requirements of operating under strict quality systems, regulatory requirements and production pressures.
• For a company building AI solutions for manufacturing operations, Pagan argues, this knowledge matters because the team is “not designing great tools in a vacuum.”
• Decisions about governance, deployment strategy and system architecture reflect the team’s understanding of the jobs to be done, the workplace environment and how decisions get made.
• Pagan also notes that their family still lives in Puerto Rico and jokes that it helps the company’s travel budget.

For brands planning a campaign

For a company seeking advisory or partnership support, this post describes a domain-credentialed team and its design philosophy, and argues who should build operations AI. The company name and team claims appear only here; ask for supporting credentials if they matter to your decision.

Treat it as context, not a source for their arguments.

Pagan briefly notes switching from manufacturing to tech at ConvergeSouth in Winston-Salem, North Carolina, thanks Peter Marsh and the Flywheel team for a strong event, and mentions Joe’s State of the South keynote. There is no argument or analysis, just a signal of attendance and participation in a regional tech gathering.

For brands planning a campaign

Useful mainly as a signal of presence in the Southeast technology circuit if a campaign targets that region or event series; it should not carry weight in assessing Pagan’s expertise.

Ignoring this variation, Pagan warns, risks supply and drug shortages.

This observation comes from one conference recap; no other post here revisits manufacturing site variability in comparable depth.

For brands planning a campaign

A talk or advisory engagement should start with site-level constraints and regulatory context, not a generic AI adoption roadmap. For a life sciences technology company, Pagan can help frame how a pilot must fit existing quality systems and production realities.

Examples: Source 1

They connect this risk to manufacturing’s existing understanding of single points of failure.

For brands planning a campaign

Position Pagan on governance and operational reliability for AI tools in regulated manufacturing. A brief can ask for concrete controls: documentation, software development lifecycle discipline, ownership, and continuity planning for AI-enabled applications.

Examples: Source 1

The next step, they argue, is knowledge-driven decision-making: combining that data with SOPs, policies, procedures, and other information that defines how manufacturing is supposed to work. They also advise against waiting for perfectly structured data before starting AI initiatives.

For brands planning a campaign

This advisory angle suits manufacturers with data integration efforts underway that need to connect them to quality and operations knowledge. A campaign can ask Pagan to map a knowledge-layer approach without overpromising on data readiness.

Examples: Source 1

They argue that the recurring gap is execution: speed to market, financial viability, regulatory compliance, and scalability determine whether innovation creates value. Pagan quotes a venture view that the best technology does not always win, pointing to execution and go-to-market.

This is a conference recap and point of view; it does not establish Pagan's investment role or endorse a specific quantum technology.

For brands planning a campaign

This is relevant to a life sciences technology company evaluating Pagan for a talk or advisory role on AI and quantum commercialisation. A session can focus on the path from discovery to regulated manufacturing scale-up, with realistic attention to compliance and market viability.

Examples: Source 1

The surrounding points frame AI as an operational capability that must fit processes, reliability expectations, and knowledge systems.

For brands planning a campaign

Position Pagan as a speaker who pushes past AI enthusiasm to operating-model questions: what problem is being solved, who owns the tool, and how it will work in a regulated plant.

Examples: Source 1

They connect that expertise to governance, deployment strategy, and system architecture, and say the goal is low barrier to entry and high usability for manufacturing users.

For brands planning a campaign

Pagan can speak to how regulatory knowledge and production experience should shape AI product decisions. A brief can ask for specific examples of how jobs to be done and workplace decision processes affect architecture and deployment.

Examples: Source 1

They note that the Puerto Rico team provides industry, regulatory, and practical manufacturing experience that shapes how the company builds.

For brands planning a campaign

Pagan can support programming on regional biopharma manufacturing capability and why location-based expertise matters for AI adoption. A life sciences company can use this angle for panels on scaling in regulated ecosystems, not just headquarters-centric innovation.

Examples: Source 1

Examples include AIMST's manufacturing AI discussions, The Economist's Commercialising Quantum, BIO 2026, and ConvergeSouth. The posts mix numbered points, personal field notes, and direct questions to the reader.

Pagan typically frames each event around a central question or tension, then uses numbered observations, contrasts between sectors, and a closing question or provocation. The posts are written in first person and often include conference-specific detail rather than generic commentary.

For brands planning a campaign

A brand can evaluate Pagan as an event-based commentator or moderator who turns conference programming into practical takeaways for manufacturing and life sciences audiences. The format suits pre-event promotion, live recaps, and post-event advisory summaries.

Examples: Source 1 · Source 2 · Source 3 · Source 4

They write about it in the first person — “our team,” “our governance, deployment strategy, and system architecture,” and a joke about how family travel affects “Murmuratto's P&L” — and cite the regulated-manufacturing backgrounds of its bio experts. Ask what their title is and whether the relationship is employment, ownership, or another commercial arrangement.

View example

View example

View example

View example

Like Pagan, they conclude that engineering, quality, automation and operations problems are rarely solved one discipline at a time.

Lewis mostly publishes industry news and deal roundups — licensing deals, regulator turnover, job cuts and trial results — and writes about integration from a project-delivery and services standpoint. Pagan's posts, by contrast, make event-based arguments about deploying AI inside a regulated plant.

Read original · Original 2

They name the same pieces that have to line up: capital, regulators, manufacturing and data.

Hemlani uses statistics to examine countries and ecosystems — sovereign wealth, FDI shares and biotech investment targets — and makes a macro-level investment case. Pagan stays at the plant and deployment level, writing from conference rooms and company work rather than analysing capital flows.

Read original · Original 2

Cooper's reported interview pieces address CIOs and CISOs across sectors, covering AI accountability, cyber resilience and trusted data. Pagan writes in the first person about manufacturing operations and life sciences, where quality systems, regulatory requirements and production uptime are binding constraints.

Read original · Original 2

Dan Lewis's posts summarise recent industry moves such as Bristol Myers Squibb's $15.2B licensing deal with Jiangsu Hengrui across 13 early-stage oncology, hematology, and immunology programmes; Takeda's plan to cut 4,500 jobs by March 2027; Eli Lilly's Mounjaro becoming the top-selling drug in Q1 2026; and the EU's provisional Critical Medicines Act deal.

• The posts also connect digital manufacturing and automation to project delivery, CQV, quality, and operational readiness, and argue that complex manufacturing challenges are rarely solved by optimising one discipline in isolation.
• This is a better fit when the audience needs a current market and lifecycle-integration briefing, or when the buyer wants a speaker who argues that engineering, quality, automation, digital systems, and operations should align around the same objectives.

The tradeoff is specificity on AI and quantum commercialisation.

• Dan Lewis's posts do not discuss quantum computing, set out a site-level AI adoption framework, or address the gap from scientific breakthrough to commercial product as directly as David Pagan's AI in Manufacturing and Commercialising Quantum posts.
• Dan Lewis's SimoTech post focuses on integrated digital manufacturing and acquisition-driven capability across the facility lifecycle, not AI-built production software, software reliability risk, or knowledge-driven operations.
• If the buyer's main ask is an AI/quantum commercialisation talk plus a workshop on regulated plant constraints, David Pagan's posts map more directly.
• Dan Lewis is the stronger alternative when the required talk is pharma market intelligence plus lifecycle project delivery.

Examples: Source 1 · Source 2

The posts describe the UAE as a coordinated biotech build-out: $2.5 trillion sitting in seven UAE institutions against only $2 billion of venture capital raised in the country in 2025; non-oil activity at 77.3% of a $517 billion GDP; Mubadala deploying $17.4 billion in the first nine months of 2025; the UAE making less than 5% of its own medicine today and aiming to become a place where medicine is invented; Mubadala holding more than $3 billion in biotech with drug discovery as its largest investment sector; a crown-prince-level Longevity Authority; a national manufacturing drive; and a population genomics programme. This is a better fit when the buyer wants an investment, sovereign-capital, and ecosystem-construction lens on biopharma manufacturing, or when the audience is deciding where to locate or partner rather than how to deploy AI on a regulated site.

The tradeoff is operational depth.

• Lavine Hemlani's posts do not cover AI deployment in manufacturing operations, quality systems, software reliability, CQV, or the day-to-day constraints of regulated production.
• They also do not discuss quantum computing.
• For a discussion of how AI and quantum move from discovery into regulated manufacturing, David Pagan's posts speak directly to site variability, AI-built tools, data-driven versus knowledge-driven operations, and the commercialisation constraints of speed to market, financial viability, regulatory compliance, and scalability.
• Lavine Hemlani is the stronger alternative when the discussion is capital formation and national industrial strategy; David Pagan is stronger when the discussion is what a biopharma site or technology buyer should do next.

Examples: Source 1 · Source 2

The posts argue that boards have moved from AI enthusiasm to demanding measurable financial and competitive outcomes; cite an August 2026 Gartner research note on evidence of business outcomes and an ISG study of 1,200 enterprise AI use cases in which 31% of prioritised initiatives reached production and only about one in four met growth-return expectations; and describe CIOs being asked to drive growth and profitability while controlling AI spend, reducing legacy technology costs, and maintaining governance.

• The second post extends the buyer task into risk: AI increases the speed, scale, and complexity of data, identities, applications, and decisions, so leaders need visibility into where information originates, how it moves, and whether its integrity can be demonstrated across preparedness, readiness, visibility, response, recovery, governance, and survivability.
• This is a better fit when the buyer's audience is enterprise technology and security leadership and the desired outcome is a board-level case for AI accountability and resilience.

The tradeoff is life-sciences manufacturing specificity.

• Lane Cooper's posts do not address biopharma site processes, quality management systems, CQV, regulatory inspections, quantum computing, or drug-discovery commercialisation.
• They focus on enterprise IT, CIO/CISO, insurance-risk, and cyber-resilience evidence.
• For a campaign about AI and quantum moving from discovery into regulated manufacturing, David Pagan's posts give the audience concrete manufacturing adoption gaps, software ownership risks, ERP/MES/QMS/PLC/SCADA-to-SOP knowledge layers, and commercialisation constraints.
• Lane Cooper is the stronger alternative when the required talk is enterprise AI governance and cyber resilience; David Pagan is stronger when the required talk is regulated manufacturing and life-sciences commercialisation.

Examples: Source 1 · Source 2