Lakshman Karumuri

Teams focus on platform selection while leaving harder transformation questions open: how much process harmonisation to expect across sites; how MES will interact with ERP, automation, lab systems and data platforms; whether to use a global template or a site-led model; and what the operating model, governance and rollout should be.

• When those choices are deferred, vendors respond to an incomplete problem statement and the real complexity shifts downstream into design, validation and scale-out.
• They recommend treating the RFP as the output of structured discovery, defining integration, data, template and governance principles first, and aligning IT, operations technology, quality and manufacturing stakeholders before vendor selection.
• That sequence, they argue, can lower total cost of ownership by avoiding fragmented deployments, clarify architectures, support right-first-time execution through integrated workflows, and strengthen validation and audit readiness through risk-based qualification, built-in audit trails and traceability by design.
• They also see MES-Lite and digital operations platforms as a way to iterate faster without losing the enterprise view.

The post lists the unresolved questions, then identifies three steps in the strongest programmes: use the RFP as the output of discovery, define integration and governance principles early, and align cross-functional teams before choosing vendors. It closes by asking what is slowing MES rollouts: technology choices, poor operating-model design or lack of enterprise alignment.

Read original

Reporting from an ISPE Pharma 4.0 community event on data as the foundation for smart operations, they highlight batch-record review moving toward review by exception with contextualised data, potentially reducing release cycles from days to hours; approaches such as knowledge graphs to bridge regulated and non-regulated data; integration layers within a unified namespace to make operations more connected and scalable; and modular manufacturing for flexibility and faster scale-up.

• They say adoption must balance innovation against total cost of ownership, return on investment, timelines and regulatory expectations.
• Pharma 4.0 differs from general Industry 4.0, they add, because compliance, validation and data integrity carry more weight.
• Their measure of success is how people work across sites and the enterprise, with people and patients at the centre of decisions.

Karumuri highlights the event's discussion of batch-record review and integration approaches, and presents the improvement in release times from days to hours as potential, not guaranteed. They credit Gareth Alford of AstraZeneca for insights on continuous API platforms and autonomous manufacturing; Kim Hewson and Hamish Sams of GSK for their discussion of integrating data across shop-floor systems; and Henal Shah for exploring the possibilities of pharma manufacturing in space.

Read original · Original 2 · Original 3

Off-the-shelf products are like tailored suits: impressive and quick to start, but not made for the buyer, so they need constant monitoring, workarounds or a full-time equivalent to keep running.

• Bespoke AI starts with a company's data and processes, making adoption feel natural and allowing the result to scale with the business.
• They point to the shift from packaged software to building on the .NET platform, arguing that the real transformation came when companies designed bespoke software around their own processes.
• In aviation, they describe the barrier as scale, not technology: moving from pilot purgatory to operational excellence.
• Language models, they argue, need to be embedded in platforms rather than bolted on.
• They also ask how chief information and digital officers should choose between AI pathways, referring to speed-to-value versus scalability matrices.

The tailoring analogy came up during a coffee conversation at 12 Hay Hill in Mayfair, where co-founder Chris Rigby gave an impromptu lecture on the difference between tailored and bespoke suits. Karumuri recounts Rigby's comparison and closes by asking readers whether they would settle for a suit that merely fits.

Read original · Original 2 · Original 3

They describe a super cycle set to surpass $120 billion by 2030, alongside a technician shortage and rising engine shop visit costs.

• Profit pools, they argue, have shifted from skilled manual labour to what they call the intelligence layer: predicting part failure with high accuracy using digital twins; using AI copilots to parse thousands of pages of airworthiness directives in seconds; connecting OEMs, airlines and maintenance providers through a digital thread into a single source of truth; building regional digital ecosystems; and cutting aircraft-on-ground risk through digitised supply chains and 3D-printed legacy parts.
• They conclude that the winners will not be the operators with the largest hangars, but those with the most integrated data—and that the super cycle's maths does not work without a radical shift in how data is used.
• They describe themself as a student of the industry, sharing their reading to pressure-test their own understanding.

From MRO Europe 2025, Karumuri relays William Gibson's line that the future is already here, just not evenly distributed, attributing it to Robert Mather of IFS. They also credit Alice Belcher of Delta Air Lines on ontology and technical debt, and report advice from Veronica Dall'Aglio of Oliver Wyman and Arne Luxa of MTU Maintenance, including Luxa's reminder to get used to failing as a route to growth.

Read original · Original 2

At an ISPE leadership summit, they expected discussion of AI and digital transformation, but found the day centred instead on purpose-led leadership, the different leadership styles needed at different stages of a journey, and self-awareness as roles, organisations and technology change: knowing what kind of leader you are, and what kind you are not.

• Their takeaway was that people were most passionate about the human questions—purpose, resilience, self-awareness and leadership—and that people remain the hard part.
• They connect this to life sciences manufacturing, where purpose means delivering medicines, vaccines and therapies that help patients, and transformation success depends on how people work across sites while keeping people and patients at the centre of decisions.
• Their account also reveals a tension: a leadership summit nearly passed without AI before the topic resurfaced late in the day, in contrast with the technology-first framing of their manufacturing posts.

Karumuri credits the sessions separately: Luisa Freitas dos Santos on purpose-led leadership in the manufacturing supply chain, Angela Bowden on leadership styles and on AI and the future workforce, and Dr Nik Kotecha on Kotecha's journey from refugee to building a global generics business — cited as an inspiring account of humility, resilience and asking for help, not as a claim about Kotecha's full biography. Karumuri points readers to Cognizant's New Work, New World 2026 research for anyone working on workforce transformation, operating models, skills and talent strategy.

Read original · Original 2 · Original 3

This post gives newcomers a clear practical answer: treat the RFP as the result of discovery, not a substitute for it.

Karumuri argues that MES programmes can lose momentum when organisations leave process harmonisation, system integration, operating models and governance unresolved until after vendor selection.

• They recommend aligning IT, operations technology, quality and manufacturing stakeholders first, then defining architecture and rollout principles.
• Questions about ERP, automation, laboratory systems and data platforms show their focus is on how an MES fits into the wider operation, not just which platform to buy.

For brands planning a campaign

A webinar on MES and AI adoption could use these unresolved decisions as its starting point. Ask participants what must be settled before an RFP, then discuss where AI and MES-Lite approaches fit without losing the enterprise view.

This post offers a concise introduction to the compliance, validation and data-integrity concerns that shape the discussion.

The post connects practical manufacturing topics—batch-record review by exception, GMP and non-GMP data, integration layers and modular manufacturing—to the realities of regulated environments.

• Karumuri emphasizes cross-site ways of working and keeping patients in view.
• The example of batch-record review describes the potential to reduce release cycles from days to hours, not a guaranteed result.

For brands planning a campaign

Use this as a basis for a webinar segment on AI adoption under regulated conditions. Ground the discussion in data integrity, validation and review workflows, and distinguish potential benefits from results that would need evidence in a specific deployment.

The tailored-suit analogy makes their argument accessible before introducing its business and technology implications.

Karumuri compares off-the-shelf AI to a suit that fits well enough but needs repeated adjustment, and bespoke AI to a solution built around an organisation’s data, workflows and problems.

• They also invoke the history of businesses building on .NET to argue that value can come from adapting technology to business needs.
• The analogy makes the case for customization, but does not establish that custom AI is always better or more cost-effective.

For brands planning a campaign

A webinar can use the analogy to open a balanced discussion about AI adoption: which needs suit standard tools, and which justify customization? Include criteria such as workflow fit, ongoing maintenance and measurable value so the session does not become a blanket pitch for bespoke solutions.

This post applies those questions to aviation maintenance rather than pharma manufacturing.

Reflecting on MRO Europe, Karumuri highlights prioritizing AI use cases by speed to value and scalability, structuring data relationships through ontology, and addressing technical debt.

• They argue that AI should be embedded in platforms rather than bolted on.
• The event-based format combines observations from named discussions with their own takeaways, showing how they turn industry conversations into technology themes.

For brands planning a campaign

For an MES-and-AI webinar, this offers a useful cross-industry comparison: invite participants to test whether the same questions about data relationships, integration and scaling apply in regulated pharma. Frame the aviation example as an analogy, not direct evidence about pharma outcomes.

This post shows a people-and-leadership perspective alongside their recurring interest in AI and transformation.

After attending an ISPE leadership summit, Karumuri says the strongest conversations focused on purpose, self-awareness and resilience rather than technology.

• They tie manufacturing purpose to delivering medicines, vaccines and therapies, then connect leadership and workforce change to AI.
• The post frames technology adoption as an organizational challenge as well as a systems question.

For brands planning a campaign

A webinar could include a short discussion of workforce readiness and leadership alongside MES and AI implementation. This would broaden the event beyond platform choices while staying grounded in the human issues Karumuri highlights.

Karumuri argues that most MES requests for proposals fail because they focus on system selection while harder questions remain open: how much process harmonisation sites need, how MES will interact with ERP, automation, lab systems and data platforms, whether the goal is a global template or a site-led model, and what the operating model, governance and rollout should be.

• They outline what getting these decisions right can unlock: lower total cost of ownership by avoiding fragmented deployments, clearer architectures, right-first-time execution through integrated workflows, and validation and audit readiness through risk-based qualification and traceability by design.
• They also present MES-Lite and digital operations platforms as ways to iterate faster without losing the enterprise view, and describe structured discovery, product evaluation, operating-model design and roadmap definition as work to complete before an RFP.

Karumuri writes from a consultancy position inside these programmes, moving between platform detail and organisational questions. They often close with a direct question, such as what is really slowing MES rollouts: technology choices, operating-model design or lack of enterprise alignment.

One MES-related post is event promotion rather than analysis: it lists platforms and named Cognizant colleagues attending Pharma MES Europe and invites meetings, but offers no argument or evidence.

For brands planning a campaign

A strong hook for a life sciences digital transformation consultancy: a webinar framed around discovery before vendor selection — how to write an MES RFP that reflects a real transformation agenda, with harmonisation, integration and operating-model decisions made first.

Examples: Source 1 · Source 2 · Source 3

In reports from pharma data and automation events, Karumuri highlights batch record review moving toward review by exception with contextualised data and cutting release cycles from days to hours; bridging GMP and non-GMP data through approaches such as knowledge graphs; integration layers inside a unified namespace to make operations connected and scalable; and modular manufacturing for flexibility.

• In aviation, the equivalent argument is that a digital thread joining OEM, airline and maintenance providers creates a single source of truth and removes manual paperwork overhead.
• They frame Pharma 4.0 as distinct from general Industry 4.0 because compliance, validation and data integrity carry more weight.

Karumuri reports specific technical approaches and named speakers from events, rather than general trends, and distinguishes regulated pharma constraints from those in other industries.

For brands planning a campaign

Useful for a webinar segment on where data work pays back first: review by exception, bridging regulated and non-regulated data, and integration layers as prerequisites for AI use cases, not follow-on projects.

Examples: Source 1 · Source 2 · Source 3

They compare off-the-shelf AI to a tailored suit that fits well enough but was never made for you, requiring constant monitoring and workarounds.

• Bespoke AI, by contrast, starts with a company's own data, workflows and pain points.
• They cite the shift from packaged software to building on the .NET platform.
• In aviation, they describe the barrier as scale rather than technology — moving from pilot purgatory to operational excellence — and point to agentic AI assistants that parse long airworthiness directives, digital twins for failure prediction, and the principle that language models need to be embedded in platforms rather than bolted on.
• They also ask how chief information and digital officers should choose AI pathways, referring to speed-to-value versus scalability matrices.

Karumuri reasons by analogy and precedent, then ties the argument to a specific delivery offer and a discovery-to-delivery sequence.

One AI post recaps a panel discussion about government, policy and the NHS, so the AI material extends beyond manufacturing audiences.

For brands planning a campaign

A webinar on AI adoption could build on Karumuri's framing: choose use cases by fit and scalability, expect monitoring and workaround costs with off-the-shelf tools, and treat bespoke development as the route to lasting adoption.

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

They describe the sector as entering a super cycle set to surpass $120 billion by 2030, while facing a technician shortage and rising engine shop costs.

• They argue that profit has shifted to what they call the intelligence layer: predicting failures with digital twins and agentic AI, using digital threads and regional autonomy, and building risk resilience through digitised supply chains and 3D-printed legacy parts to reduce aircraft-on-ground costs.
• Their conference reports cover maintenance software, immersive tools, technical debt, ontology and why early adopters lead.

Karumuri reports from events as a participant, naming speakers, vendors and counterparties, and often closes with a reading list of related topics.

One aviation post is a short attendance notice with a registration link, not analysis.

For brands planning a campaign

This beat matters mainly for brands spanning industrial, aerospace or enterprise technology. It shows they can carry the same argument — that data and integration determine outcomes — across two regulated, asset-heavy industries, supporting a cross-industry or comparative webinar format.

Examples: Source 1 · Source 2 · Source 3

At a life sciences leadership summit, Karumuri expected AI talk but found the day centred on purpose-led leadership, leadership styles at different stages of a journey, self-awareness as roles and organisations change, and a speaker's journey from refugee to building a global generics business.

• Their conclusion: the conversations people cared about most were the human ones — purpose, resilience, self-awareness and leadership — and people remain the hard part.
• Related notes include advice to get used to failing as a route to growth and an emphasis on keeping people and patients at the centre of transformation decisions.

Karumuri builds arguments from named sessions and personal reactions at events, then states a takeaway plainly.

In that post, they note that a full leadership summit nearly passed without discussing AI — the opposite of the technology-first framing in their manufacturing posts.

For brands planning a campaign

Offers a consultancy a second webinar angle beyond the technical: workforce transformation, skills and operating-model change in the age of AI. That suits a life sciences audience where adoption depends on site-level change.

Examples: Source 1 · Source 2 · Source 3

When Karumuri makes an argument rather than reporting, they use numbered frameworks — such as four shifts in aviation maintenance, or three practices that distinguish the strongest MES programmes.

• Longer explainers work through a single analogy, and engagement tends to come through direct questions to readers or requests for thoughts.
• They share industry research and event registration links when relevant, and tag the employer and practice groups they represent in manufacturing posts.

Conference recaps with named attribution; numbered lists; analogy-led explainers; closing questions to the audience.

Some posts are scheduling or attendance notices with little beyond logistics; a selection of Karumuri's strongest posts offers a better sample of their work than the full feed.

For brands planning a campaign

Karumuri suits a host or panelist role more than a scripted keynote: they summarise other people's sessions accurately, give credit and turn a day of talks into a few usable takeaways. That fits a thought leadership webinar with multiple speakers.

Examples: Source 1 · Source 2 · Source 3 · Source 4 · Source 5 · Source 6 · Source 7

Their posts present Cognizant as their employer and vantage point for pharma and manufacturing analysis.

View example

They also hosted a Colab-8TalksAI panel with former MP James Morris and tagged both Colab-8 Technologies and Colab-8 Consulting. Ask for their title and ownership share.

View example

They attended the ISPE Leadership Summit at GSK's headquarters and the ISPE UK Affiliate Pharma 4.0 session at AVEVA London, writing both times as a Cognizant representative, and thanked the ISPE UK Affiliate for hosting.

View example

Both put alignment ahead of tooling and write in the first person about what they see going wrong in client programmes.

Hier's material covers business roadmapping, prioritisation grids and the visual collaboration software they represent, staying at the level of strategy and portfolio decisions. Karumuri gets into manufacturing systems, integration layers and AI fit on the plant floor and in maintenance operations.

Read original · Original 2

Both write about how connected the stages of a facility's life have become, from design through commissioning and operations. Readers following one for industry manufacturing news may find the other useful on the same subjects.

Lewis posts mostly as a news and announcement channel, with roundups of pharma industry deals and regulatory moves, plus a completion notice for an acquisition that adds automation and digital manufacturing capability. Karumuri writes analytical posts that set out a framework or argument and invite a response.

Read original · Original 2

Lewis's work on regional vaccine manufacturing and supply chain resilience and Karumuri's work on operating models and data across sites share a focus on how these systems are organised rather than on individual products.

Lewis focuses on supply chain strategy, crisis response and equitable access to medicines at a policy and programme level. Karumuri focuses on manufacturing execution, plant data and AI adoption on site, as well as aviation maintenance alongside pharma.

Read original · Original 2

Working the same buyer task — helping a life sciences organisation get research, regulatory, manufacturing and commercial plans moving together — Hier offers a mechanism, not just a diagnosis: collaborative visual roadmaps that surface dependencies, gaps and bottlenecks before they become costly, and a prioritisation exercise where stakeholders co-create the must-haves so buy-in exists before budgets are set.

• If the consultancy sells roadmap definition and operating-model alignment to life sciences leaders, Hier can run an exercise the audience can repeat the following week.
• The posts also carry a candid, self-critical note that is uncommon in this space: they write about coaching clients on alignment while admitting they keep working in their own business instead of on it, and about updating their own roadmap for the year ahead.
• That honesty is useful for a session that has to persuade sceptical manufacturing and quality leaders rather than impress them.

The tradeoff is distance from the webinar topics and a commercial tie.

• Neither post touches MES selection, integration with ERP, automation or lab systems, batch record review, or AI adoption — the topics the webinar names.
• The basis is two posts across roughly five months, both persuasive arguments rather than field reporting from events, so there is no track record of summarising other speakers' sessions or moderating a multi-speaker panel.
• They write as a practitioner selling roadmapping work delivered through a named software platform, so a webinar built on their framing can read as a product demonstration unless the format deliberately separates method from tool.
• Ask for audience composition and reach.

Examples: Source 1 · Source 2

Their news roundup works through deal flow, regulatory disruption, manufacturing restructuring and AI drug-design funding in a single pass — a large oncology licensing deal, leadership churn at a medicines regulator, thousands of jobs cut in a restructuring drive, shifts in the world's top-selling drug, EU measures to secure supply of essential medicines, a multi-billion raise for an AI drug-design platform — which suits an executive audience that needs commercial and policy context before plant-level discussion.

• Their second post is closer still to the webinar's focus on MES and AI adoption in pharma manufacturing: the completion of an automation and digital manufacturing acquisition, argued through the whole facility lifecycle from design and project execution through commissioning, qualification and validation, start-up, operational readiness and ongoing operations.
• The stated conclusion is the one the consultancy wants on the table: the hardest manufacturing problems are not solved by optimising one discipline in isolation, and the best outcomes come when engineering, quality, automation, digital systems and operations align on the same objectives.

The tradeoff is depth on the webinar's decision about MES and AI adoption in pharma manufacturing.

• Two posts is a thin base, and neither addresses RFP framing, process harmonisation across sites, global template versus site-led models, integration architecture, or the reasons AI pilots stall at scale.
• The roundup scans headlines rather than analysing them, so the audience gets awareness without method.
• The acquisition post is corporate news commentary tied to their employer's deal, useful for the lifecycle argument but not neutral.
• Ask for audience data, event experience, commercial terms and a booking route.

Examples: Source 1 · Source 2

One post takes pharma and med-tech supply chains as the problem and describes a deliberate move from business-as-usual to crisis response, with a framework used to deploy capabilities to client challenges, validate existing responses and prepare scenarios for an uncertain future.

• The other sets out approaches regional stakeholders — global and regional manufacturers, government agencies, NGOs and financiers — could take to advance regional vaccine manufacturing, co-authored with colleagues and framed around equitable access and health security.
• For a webinar on supply security, regional manufacturing or cross-sector collaboration, that is a closer subject match than Karumuri's event reporting and platform-adoption arguments.

The tradeoff is currency and specificity.

• The posts date from 2020 and 2024, so the assessment rests on older material, with no recent activity to inspect; that limits competitor comparison and currency checks.
• Nothing in them covers MES, AI adoption, data integration, batch record review or the operating-model questions at the centre of a webinar on MES and AI adoption in pharma manufacturing.
• Neither post is first-person field reporting: both are firm perspective pieces, and the substantive detail sits in linked publications rather than in the posts themselves, so a webinar would inherit claims the posts cannot support in depth.
• Ask for hosting experience, audience data and commercial terms.

Examples: Source 1 · Source 2