Judgment First: Subhomoy Bakshi and the Making of a Future-Ready Law Firm 

https://www.lkslaw.com/Subhomoy Bakshi is Head of Digital Transformation at Lakshmikumaran & Sridharan Attorneys, the firm known simply as LKS, a full-service practice with more than 550 fee earners across 14 offices. An alumnus of the Indian Institute of Technology, Kharagpur, where he earned his B.Tech (Hons) and M.Tech in Mechanical Engineering, he has built a career across manufacturing, supply chain, education, and legal technology, unified by a single thread: transforming legacy systems so that they serve more people. 

That thread began at ITC Limited. There, he worked on the digital transformation of a 120-year-old factory, managed capital expenditure projects for his site, and was part of a team that kept one of India’s largest supply chains running when every system collapsed during COVID.  

A training system design project that he was a part of went on to win an international excellence award. He then founded an education venture that carried digital transformation into underfunded public and private schools across four Indian states, taking quality learning outcomes to students the system had largely overlooked. 

The Mature Capability 

Today, from New Delhi, Subhomoy builds AI-powered legal products designed around a simple conviction: technology should deepen the quality of counsel, never dilute it. The firm commissioned India’s first known in-house law firm GPU infrastructure for proprietary model training, and its AI-powered legal intelligence platform, built on more than four decades of institutional knowledge, was deployed and fine-tuned over a disciplined nine-month cycle before firmwide launch. The maturity shows in the usage: close to 10,000 AI-assisted research interactions and more than 2,500 citation intelligence lookups within the first three months, with research outputs operating at 93 to 99 percent accuracy and every answer traceable to source. The work has drawn notice beyond the profession as well: Dell Technologies featured the firm’s agentic AI platform in a global customer story, Delivering Legal Expertise at Scale with Agentic AI, showcased at its worldwide summit. 

None of this is automation for its own sake. Every system is built so that lawyers spend less of their day hunting for information and more of it exercising judgment on that information; what the client receives is deeper, more consistent, better-evidenced counsel from the same experienced hands. That is what a mature technology capability looks like in professional services: measured in accuracy, provenance, and adoption, not in announcements. 

The Grounded Foundations 

Along with a career journey that spans entrepreneurship, AI strategy, product leadership, and digital transformation, Subhomoy learned leadership on a factory floor before he learned it in a boardroom. His early years were spent in heavy manufacturing, running shifts alongside teams of hundreds, many of whom had decades more experience than he did and far less formal schooling. You cannot fake credibility in that setting. He earns it by being useful, by listening first, and by doing the unglamorous work himself. One product from those years, a training system that he designed, went on to win an international excellence award, but the deeper reward was learning how change takes hold in people who have no reason to trust him yet. Founding an education company afterward reshaped the rest. Building from nothing, with no permission to fail and no budget to waste, teaches a bias for outcomes over activity. Working close to AI products taught him to respect both their power and their limits. The factory, the startup, and the AI lab left him with one conviction: technology is the easy part; judgment and adoption are the hard part. 

The Frictionless Practice 

As the Head of Digital Transformation at LKS, his vision is a firm where technology removes friction so that lawyers can do more of what only lawyers can do: think, judge, and counsel. Legal work is knowledge work at its purest, and knowledge work is where AI offers the most leverage and the most risk. Future-ready, to him, does not mean automating lawyers out of the loop. It means building tools that make every practitioner faster at research, drafting, and review, while keeping professional judgment firmly in human hands. A future-ready legal organization treats its data as an asset, grounds every tool in its own trusted knowledge rather than the open internet, and earns adoption rather than mandating it. It also opens that knowledge outward. They now publish independent knowledge portals on complex, fast-moving areas of law, from GST to employment law to the EU’s Carbon Border Adjustment Mechanism to sustainability and ESG, as a contribution to the public understanding of these fields. The portals are the firm’s oldest instincts made public: knowledge pursued with passion, held to a standard of integrity, and shared with anyone who cares to learn. The goal is not a more digital firm for its own sake. It is a firm that serves clients with greater speed, consistency, and confidence, built on an infrastructure they trust completely. 

The Cultural Shift 

Digital transformation is often viewed as a technology initiative. But for him, technology is the trigger, never the transformation. A new tool only matters when people change how they work around it, and that is a business and cultural shift, not a technical one. So he treats every rollout as a change program first and a software project second. The build is maybe a quarter of the effort; the rest is adoption. They make it cultural by starting from real work, not features. They anchor training in people’s live matters, run open houses where skepticism is welcome, celebrate early adopters openly, and watch engagement signals closely; a falling feedback rate tells them a tool has not earned its place yet. Culture does not change in a town hall; it changes one convinced user at a time. At LKS, that patience shows in the numbers: 87.5 percent of the firm’s practitioners now use its AI tools as part of their everyday work. When usage becomes a habit that people defend, the transformation has happened. Until then, all you have is software. 

The Operating Discipline 

According to Subhomoy, the biggest opportunities and challenges of integrating AI into operations today are two sides of the same coin. The opportunity is enormous: AI can compress the time spent on repetitive knowledge tasks, freeing skilled people for higher-value work. The challenge is that most organizations underestimate everything that makes that opportunity real. The first challenge is data. AI is only as good as the corpus it is grounded in, and preparing that corpus well is consistently underestimated. The second is trust and governance, especially where confidential information is involved; for them, that means keeping data on infrastructure they control rather than sending it to external systems. The third is hype. It is easy to buy tools and announce progress; it is hard to drive real adoption and measure real value. The organizations that win treat AI as an operating discipline rather than a press release, and that discipline is what separates a tool that changes how people work from one that gathers dust after a launch event. 

The Disciplined Innovator 

Having worked across startups and established enterprises, Subhomoy’s approach to innovation evolved over the years. Early on, he equated innovation with building. If something was broken, he wanted to build the fix himself. Over time, and especially after running a company where every rupee and every hour counted, he learned that the smartest innovation is often knowing what not to build. Today, his approach is disciplined rather than enthusiastic. They make build, buy, or hybrid decisions through a clear scoring framework rather than instinct, weighing strategic value, risk, and control, and how feasible execution truly is. Startups taught him speed and the courage to ship; enterprises taught him the patience to make change stick at scale. The synthesis is a kind of impatient rigor: move quickly, but only on the things that deserve to be built, and never confuse motion with progress. 

The Paced Transformation 

He balances technological advancement with the human side of change management during transformation initiatives by sequencing rather than forcing. Not every task should be automated, and pretending otherwise is how you lose people’s trust. Forcing automation where people are not ready does not speed transformation; it stalls it. He uses a simple framework that sorts work by two axes: how human-intensive it is and how intelligence-intensive it is. That tells them what to automate now, what to merely augment, what to keep human-first, and what to defer. The human side gets the heavier weighting on purpose. People do not resist technology; they resist feeling replaced or rushed. So they move fastest where machines clearly help, and people are glad to hand the task over, and they move slowly, with consent and training, where judgment and relationships live. Advancement that respects that line is adopted. Advancement that ignores it is resented, and resented tools do not get used. 

The Sharper Professional 

Subhomoy believes data, automation, and artificial intelligence play a crucial role in reshaping professional services and knowledge-driven industries. In professional services, the product is judgment, and judgment runs on information. Data, automation, and AI are reshaping how quickly and reliably information reaches the professional. A researcher who once spent hours finding relevant precedent can surface it in moments, then spend the reclaimed time on analysis and counsel. But knowledge work has a non-negotiable principle: provenance. In law, especially, an answer is worthless, even dangerous, if you cannot trace where it came from. So every tool they build is grounded strictly in trusted sources, refuses to invent, and keeps a human reviewer accountable for the final word. Human-in-the-loop is not a safety caveat they bolt on; it is an architectural principle they design around from the start. Done this way, AI does not replace the professional. It makes the professional sharper. 

The Trust Framework 

He has spoken about learning from mistakes and sharing those lessons with others. Failure has contributed to his growth as a leader. His most useful failure came early. He tried to drive a digital transformation inside a 120-year-old factory, working with people who had spent their lives mastering a craft, many of whom had left formal schooling around the sixth standard. He arrived with good technology and an assumption that it would speak for itself. It did not. The workforce had no reason to trust a new system and no appetite to change a way of working that had served them for decades. He had built the easy part and ignored the hard part. That experience taught him the lesson he now builds everything around: change management is at least three times more difficult than the technology itself. People do not resist change because they are slow; they resist because change asks them to risk what they are good at. Earn that trust first, and the technology follows. Skip it, and nothing ships. 

The Execution Advantage 

When it comes to the emerging digital trends Subhomoy believes will have the greatest impact on businesses over the next five years, three stand out. First, agentic AI: systems that do not just answer questions but carry out multi-step work under oversight. This will reshape operations far more than chat interfaces have. Second, the shift toward private, grounded AI: as data sensitivity becomes a board-level concern, more organizations will run capable models on infrastructure they own, close to their own knowledge, rather than relying solely on public services. Third, and most underrated, the professionalization of adoption itself. The differentiator over the next five years will not be access to AI, which everyone will have, but the organizational muscle to embed it in practice: governance, training, measurement, and culture. The technology is converging. The advantage will come from execution. These are questions he also pursues in public. His ongoing series of sixty essays, What Intelligence Requires, maps the structural constraints on general intelligence through five lenses at once: AI systems, anthropology, law, economics, and science fiction. 

The Practical Advice 

The advice Subhomoy would like to give to aspiring digital transformation professionals who want to lead large-scale change initiatives is straightforward. Fall in love with the problem and the people, not the technology. The tools will keep changing; the discipline of understanding a workflow deeply and earning the right to change it will not. He offers three practical things. First, spend disproportionate time on adoption, because that is where value is realized, not in shipping the tool. Second, be rigorous about what to build versus what to buy, and resist the urge to construct everything yourself. Third, stay close to the ground. The best transformation leaders he knows move fluently between a server room and a shop floor, between a strategy deck and a frustrated user. Credibility is built in that range. Lead with usefulness, measure honestly, and let results, not announcements, do your talking. 

The Defining Qualities 

In his opinion, for an inspiring digital transformation leader in 2026, the defining quality is range paired with restraint: enough technical depth to make real decisions, and enough discipline to keep humans and judgment at the center. Beyond that, curiosity, humility, and real respect for the people whose work you are changing matter more than any credential. The best transformation leaders he knows are bilingual, fluent with engineers and with skeptical end users alike, and comfortable being the least impressive person in a room of experts. They are not the loudest voices about AI; they are the ones quietly making it useful, day after day. Above all, they are trusted because they have earned it by being useful before being visionary. 

The Cultivating Impact 

As one of Business Today Global’s ‘Most Inspiring Digital Transformation Leaders – 2026,’ Subhomoy believes that transformation is not a technology you purchase; it is a behavior you cultivate. The build is the smallest part of the work. Lasting impact comes from trust, patience, and the unglamorous discipline of helping people change. Ground everything in knowledge you can trust, keep judgment in human hands, and respect the people whose working lives you are asking to change. Get the human side right, and the technology will more than repay you. If he could leave leaders with one line, it would be this: build less, listen more, and protect the trust of the people doing the work. That is not the slower path to impact. Over time, it is the only one that lasts.