Thirty-Five Years Inside the Reactor: What Dr. Sambireddy Kurre Learned Running India’s Pharmaceutical Engine
He started as a chemist processing two tonnes of drug per month. He finished overseeing operations across multiple locations producing at a scale that required continuous investment, 24×7 shift coverage, and regulatory compliance across three global pharmacopoeias. The lessons in between have little to do with inspiration and everything to do with how complex operations actually function.
Dr. Sambireddy Kurre completed his chemistry graduation in 1989. His first job was at a private company in the Dr. Reddy’s Group, working as a manufacturing chemist on a single product, producing two tonnes per month. He retired from the same geography — Visakhapatnam, Andhra Pradesh — as head of operations for two sites of a pharmaceutical manufacturer, overseeing production volumes many times what he first managed.
In between, he navigated three corporate structures at the same physical location, led a startup acquisition that failed not because of operational problems but because of capital structure, earned a PhD in environmental sciences while working full time, and watched the industry transform from keyboards and hand-written batch records to ERP systems and nascent Industry 4.0 digitalisation.
His interview for The Grey Space offers something valuable to leaders who manage asset-heavy, people-intensive operations: a detailed account of how complex manufacturing decisions are actually made at each level of seniority, how technology adoption moves through a regulated industry, and where the structural vulnerabilities in India’s pharmaceutical manufacturing base sit today.
The Chemistry of Operations
The pharmaceutical API manufacturing process is, as Dr. Sambireddy explained it, fundamentally a form of controlled cooking at industrial scale. Chemicals are added to reactors in defined ratios, heated to specified temperatures, held for defined durations, and the desired compound is extracted in its purest form. The impurities are separated and treated as waste.
The difference from domestic cooking is that the recipes are governed by global regulatory agencies — the United States Pharmacopeia, the Indian Pharmacopeia, the British Pharmacopeia — and that the cooking happens in reactors ranging from 100 litres to 25,000 litres, running continuously for a week at a time, across three shifts, around the clock.
Every batch in this process involves in-process quality control at each critical step. A sample is submitted to the quality control laboratory. The batch waits. The result comes back. If it meets specification, the next step proceeds. If it does not, the batch either requires remediation or is rejected. This is not a periodic audit process — it is the operational architecture of every production run.
“Without quality control, nothing happened,” Dr. Sambireddy said. “Without quality control, you cannot move the batch forward.”
For leaders who manage operations in less tightly regulated industries, this is a useful reference point. Quality control in pharmaceutical manufacturing is not a department that checks what operations produced. It is the gate through which every production decision must pass. Removing it, compressing it, or treating it as an overhead cost is not efficiency — it is a failure of operational governance.
The Decision Protocol at Each Career Level
What makes Dr. Sambireddy’s account analytically useful is his description of how decision-making changes as career level changes.
At the chemist level, decisions are straightforward: follow the batch manufacturing record, execute the process steps, submit in-process samples, record observations. The judgment involved is limited.
At the block in-charge level — managing 20 to 30 people across around-the-clock operations — the decisions are different. When a chemist on the night shift encounters an unexpected result in a live reaction, they need an immediate direction: continue, stop, adjust. The in-charge is available at any hour, carries the historical batch data in memory, and provides the call.
At that stage, decisions were based on “historical experience and data from previous batches,” Dr. Sambireddy explained. When data was insufficient, the response was to run a laboratory-scale experiment first, observe the behaviour, and then proceed at commercial scale. No step was skipped under production pressure — the hazard stakes were too high.
This protocol — historical data, then controlled experiment, then commercial action — is a decision model that leaders in any operations-intensive business should examine. It is not conservative; it is risk-managed. The alternative, proceeding based on intuition alone under production pressure, is how process industries generate their worst incidents.
At the cluster head level, the operating model shifts again. Now the question is not whether this batch should proceed, but whether the entire monthly production plan will deliver the budgeted revenue. The decisions are about capacity allocation, product scheduling — runners versus campaigns — procurement lead times, and overhead absorption.
The key insight here is that runners and campaign products are fundamentally different cost structures. A runner product is manufactured continuously, so equipment utilisation is high and overhead per unit is low. A campaign product — manufactured once or twice a year — sits on idle equipment between runs, and its overhead absorption is therefore much higher per unit. Planning the right mix of runners and campaigns, and sequencing them correctly through available reactor capacity, is where the operations head creates or destroys margin.
The Startup That Taught Capital Discipline
Between the large corporate career stages, Dr. Sambireddy spent time as operations head of a startup — an acquisition-based pharma venture with a US-based promoter. The plan was to manufacture both API and dosage forms. The capex came from the promoter; working capital was to come from financial institutions.
The financial institutions, as they routinely do, required operating history before releasing significant credit. The promoter ran into his own business difficulties in the US and could not supplement the capital. The available working capital was sufficient for dosage form operations only. The API business — which was Dr. Sambireddy’s domain — could not be funded, and he returned to his original company.
The business diagnosis is precise: the failure was not operational. The facility worked. The people were competent. The products were commercially viable. The failure was capital structure. A single-source promoter model with no independent working capital facility created a binary dependency that, when disrupted, left no recovery path.
This is a more common failure pattern in manufacturing businesses than the startup mythology of product-market fit would suggest. Plants can be operationally excellent and financially insolvent simultaneously. The capital structure and the operational structure must both be tested independently before committing to scale.
Technology: The Layer That Is Still Being Built
The industry Dr. Sambireddy describes is in an uneven digital transition. ERP systems — SAP and Oracle — arrived in the early 2000s and brought genuine planning capability to an industry that had previously used manual tables, Xerox copies, and tick marks to manage production scheduling. The step change was real.
What followed has been less coherent. The industry is beginning to digitalise laboratory management systems, warehouse management, manufacturing execution, and EHS tracking — but the pace is described as slow, constrained by two factors that Dr. Sambireddy articulated clearly.
The first is hardware. Legacy facilities were designed for manual operation. The physical infrastructure — reactors, piping configurations, sensor placements — was not built to support the data capture that modern MES or process analytics systems require. Retrofitting software on top of hardware that was not designed for it generates exactly the pattern he described: a vendor who recommends one server, then identifies a conflict requiring a second, then proposes a historian server to manage both. The cumulative cost of the upgrade exceeds what was budgeted, and the result is a system that still does not deliver the intended capability.
The second constraint is people. Digital systems require people who can operate them. In an industry that has historically recruited chemists and process engineers, building a population of employees who are comfortable with data systems, software interfaces, and digital record-keeping is a change management challenge that takes years, not months.
His recommendation was direct: for new facilities, build digital from the start. A greenfield plant can be designed with the sensor infrastructure, the data architecture, and the network topology that makes digitalisation native rather than retrofitted. The cost comparison that matters is not the cost of digitalising an old plant versus doing nothing — it is the cost of digitalising an old plant versus building a new one correctly.
What the Industry Needs Next
Dr. Sambireddy’s forward view is structured around a clear assessment of what India’s pharmaceutical industry has been — and what it must become.
It has been, for thirty years, an excellent manufacturer of off-patent molecules for global markets. The currency advantage is real. The talent pool is genuine. The scale achieved at Hyderabad, Visakhapatnam, and Gujarat clusters is internationally significant. These are not small achievements.
But margins are compressed. The off-patent model has commoditised. The players are many. The pricing power resides with customers, not suppliers.
The next layer of the value chain — biologics, biosimilars, precision oncology, RDNA-based therapeutics — requires different infrastructure, different capabilities, and a much higher regulatory bar. The manufacturers who will capture that value are those who are making the infrastructure investment now, not when the competitive pressure is already acute.
At the same time, Germany and other European countries are actively recruiting Indian pharmaceutical talent to fill their own skill gaps. This is both an opportunity and a warning: India’s talent export advantage is real, but if the domestic industry does not build the facilities, systems, and products to retain and utilise that talent at home, the export will accelerate.
Conclusion
Dr. Sambireddy Kurre’s career is a study in operational depth. He did not diversify early. He did not pivot to higher-margin segments. He went deeper — into the chemistry, the systems, the regulatory frameworks, the people structures that make large-scale pharmaceutical manufacturing function.
The result is a body of knowledge that no amount of formal qualification can replicate: the contextual judgment to know, at two in the morning, whether a batch should proceed or be stopped; the financial literacy to understand why a well-run plant shut down; the institutional knowledge of three corporate structures, an ERP implementation, and a digitalisation journey — all from the same operational ground.
For leaders managing asset-heavy, people-intensive operations, the lesson is not inspirational. It is practical. Build the decision protocol into the system. Test the capital structure independently of the operational plan. Do not conflate technology procurement with technology strategy. And understand that the industry’s obligation to quality is not negotiable — because the person at the end of the supply chain has no alternative to receiving what you ship.
