Five roles, from the control room to the risk committee

Mandate

Make the plants' computerised systems inspection-ready and recoverable, prepare IT controls for a listing, and give customers usable data without exposing one customer's process to another.

Outcomes

• Cut batch-record review at Dahej from 6.5 days to 1.8 days for digitised products, with documentation errors down from 31 to 4 per 100 batches.
• Closed 11 IT general control deficiencies before the November 2023 listing; the auditors' report on internal financial controls for 2022-23 was unmodified.
• Kept both plants running through the February 2025 ransomware attack; ERP was rebuilt cleanly in 11 days, and restored in 58 hours in the November 2025 drill.
• Built a segregated customer data platform serving 31 customers, cutting batch-data turnaround from nine days to two.

Decisions in context

The board appointed me chief technology officer in April 2021, bringing corporate IT, plant systems, laboratory platforms and information security under one executive as Trivikaya prepared to list. My first two years were dominated by that preparation: the IT general controls our auditors had raised, the technology risks described in the offer document and the Dahej execution-system programme that customers had begun to require. Since the November 2023 listing I have sat as an executive member of the Risk Management Committee and report each year to the Audit Committee. The February 2025 ransomware attack tested the segmentation decisions of 2017 and 2019; the plants ran, the rebuild took eleven days and the drills since have brought that to 58 hours. I lead 212 people through seven direct reports, and I work most closely with the heads of quality and manufacturing, whose decisions my systems exist to support.

Reason for the move

Current role. Not seeking an executive move before the Kurkumbh programme completes in late 2027; considering one independent directorship with the board's consent.

Mandate

Replace a founder-era patchwork of standalone applications after the private-equity acquisition, and bring the plant networks under control.

Outcomes

• Moved every site onto a single ERP instance in 2018.
• Consolidated 23 standalone laboratory and plant applications into one laboratory information system and a site historian at Kurkumbh.
• Separated the Kurkumbh control networks from corporate IT in 2017 and designed Dahej's networks segmented from the start in 2019.
• Kept the research centre and both plants running through the 2020 lockdowns, with remote access for review roles approved by quality.

Decisions in context

I joined Trivikaya in September 2016, a few weeks after Corvanta Capital Partners bought control from the founding family. The company had ₹690 crore of revenue, one plant at Kurkumbh and twenty-three standalone applications, several of them understood by a single person. I reported first to the CFO and from 2018 to the managing director. We moved every site onto one ERP in 2018, consolidated the laboratory and plant applications into a single laboratory information system and a site historian, and separated the plant control networks from corporate IT, a project several colleagues considered expensive at the time. When Dahej was designed in 2019 its networks were segmented from the first drawing. During the 2020 lockdowns we kept the research centre and both plants working with remote access for review roles, each approved by quality. By March 2021 my team had grown to 96 and the company's revenue to ₹1,180 crore.

Reason for the move

Promoted to CTO when the board combined IT, plant systems, laboratory platforms and information security under one executive ahead of the listing.

Mandate

Introduce a manufacturing execution system and a common laboratory information system at the company's two USFDA-inspected API plants.

Outcomes

• Deployed the company's first electronic batch records at Ankleshwar for 22 APIs.
• Rolled out one laboratory information system across three quality-control laboratories, replacing data held on individual instrument workstations with networked capture.
• Presented the data-integrity programme to the Audit Committee in 2015.

Decisions in context

Saptadhara Organics, a listed API manufacturer in Vadodara, hired me in June 2011 to lead manufacturing systems at its two USFDA-inspected plants at Ankleshwar and Panoli. I reported to the vice president of operations, with a functional line to the CIO, and ran a team that grew to 28. We introduced the company's first manufacturing execution system at Ankleshwar for 22 APIs and a single laboratory information system across three quality-control laboratories, replacing data held on individual instrument workstations with networked capture. In 2015 I presented the data-integrity programme to the Audit Committee, which taught me to explain a programme by what could go wrong rather than by its features. Five years inside the plants showed me what a vendor rarely sees: which controls survived contact with the shift and which were quietly bypassed. I left because corporate IT and security sat in a separate function, and I wanted responsibility for all of it.

Reason for the move

Wanted end-to-end responsibility for a company's technology, including corporate IT and security, which Saptadhara kept in separate functions.

Mandate

Engineer and commission batch-control and plant automation systems for pharmaceutical and fine-chemical manufacturers.

Outcomes

• Commissioned control and batch-automation systems at 17 plants, 11 of them pharmaceutical API units.
• Wrote the company's first validation templates for pharmaceutical customers, aligned with their qualification protocols.
• Led alarm rationalisation on two retrofits, cutting alarms per operator by about 70 per cent.

Decisions in context

I joined Kiyosaka Controls India in January 2006 at its Vadodara project centre, as a project engineer and from 2008 as project manager for life-sciences projects. The company was the Indian subsidiary of a fictional Japanese automation group and sold distributed control and batch-automation systems to chemical and pharmaceutical plants. I engineered and commissioned systems at seventeen plants, eleven of them API units, leading teams of six to fourteen engineers on contracts worth ₹4 crore to ₹26 crore. Pharmaceutical customers expected documented evidence that each system worked as specified, and I wrote the company's first validation templates for them. On two retrofits I led alarm rationalisation that cut alarms per operator by about 70 per cent. I completed a part-time master's degree in computer science in 2010. I left because a vendor rarely learns whether its systems are still used as designed three years after handover.

Reason for the move

Wanted to stay with one company's plants long enough to see whether the systems were still used as designed three years later.

Mandate

Maintain field instruments and the plant's distributed control system, and later lead the control-system migration of the membrane cell house.

Outcomes

• Led the 2004 move of the membrane cell house from relay panels to the distributed control system during a 19-day shutdown.
• Ran a three-year alarm review after a 1999 chlorine-header upset, cutting standing alarms from about 1,100 to 180.
• Introduced calibration records tied to each instrument's tag number.

Decisions in context

I joined Ushmavati Chemicals in July 1998, straight from engineering college, as an instrumentation engineer at its integrated chlor-alkali complex near Bharuch. I maintained field instruments and analysers, then the plant's distributed control system, and by 2005 led a shift crew of nine as senior engineer for control systems. A chlorine-header upset in November 1999 that produced more than four hundred alarms in ten minutes started a three-year alarm review, which cut standing alarms from about 1,100 to 180. In 2004 I led the move of the membrane cell house from relay panels to the control system during a nineteen-day shutdown, and I introduced calibration records tied to each instrument's tag. Seven years in one plant taught me how operators actually use a system under pressure, and how quickly they stop trusting one that cries wolf. I left to see how control systems were designed rather than only maintained.

Reason for the move

Joined an automation vendor to see how control systems were designed rather than only maintained.