Show Notes
In this episode of Software People Stories, I speak with Mario Lewis, a manufacturing specialist turned IT expert with over 35 years of experience. He talks about his career that began far away from software—in mechanical engineering, manufacturing, operations, and quality—and eventually found its way into the software services world. We talk about what manufacturing can teach software teams, why domain knowledge and business understanding matter, how to stay adaptable through career transitions, and Mario’s experiment with AI-assisted software development through Project Klyve. It’s a thoughtful conversation about learning, reinvention, and doing meaningful work—even when the path is unexpected.
Highlights from this conversation:
Early passion for aviation led to mechanical engineering, manufacturing, and operations research. Manufacturing taught lessons in process rigor, quality, planning, and business constraints. Software became a chosen path, not a passion, but Mario approached it with discipline and curiosity. Adaptability emerges as a key career skill across industries and eras of change. Mario argues that technology is an enabler; real value comes from solving business problems. Domain depth and tacit knowledge are becoming more important, especially in GCC-style models. Project Klyve tested how far AI and agentic tools could go in building a complete application. Advice for early-career professionals: master fundamentals, understand hardware, networks, data, and security. Advice for mid-career professionals: keep learning, build domain depth, and understand market forces. Retirement feels comfortable because IT was meaningful work, but not Mario’s core passion.
Mario began his career on the shop floor with planning and manufacturing optimization roles in machine tool and switchgear manufacturing. That early focus on physical manufacturing and basic operations grounded his transition into IT services and enterprise software, where he spent over 35 years leading delivery, consulting, operations and business teams. Along the way, in 2006, he authored a book on the realities of IT service offshoring based on his practical experience in the field.
At the end of 2024, Mario stepped away from full-time corporate roles into retirement. He spent that time building Klyve, an automated desktop software factory that implemented an SDLC to build full software applications using agentic AI. It was built around strict engineering controls, deterministic design, and local data privacy.
For Mario, managing career shifts comes down to fundamentals. Staying useful across changing industries requires an openness to lifelong learning, disciplined routine, steady execution, and personal standards while tracking industry and technology trends."
Mario may be reached at: mariolewis @ gmail.com.
The Klyve repo is at https://github.com/mariolewis/klyve_factory
His book (from 2006) is available at https://www.amazon.in/Application-Service-Offshoring-Insiders-Guide/dp/0761935258
More from Software People Stories
“AI gives you wings,” “Wrong things in, rogue things out,” and “Each one, teach one.”
From democratizing intelligence to redefining how we lead, work and remain human In this episode of Software People Stories, Gayatri Kalyanaraman reconnects with Zafar, who brings nearly four and a half decades of experience across technology and multiple waves of transformation—from mainframes and personal computing to the internet, cloud and digital transformation. His observation about AI is simple but profound: previous technology shifts gave humans more powerful tools; AI is beginning to democratize intelligence. And with AI agents, even execution is being democratized. Beyond AI adoption: where is the business value? Zafar is enthusiastic about AI's potential, but cautious about the hype. Giving thousands of employees access to AI tools is not, by itself, transformation. The real question is: what has changed in the way the organization creates value? For him, meaningful enterprise value will come from redesigning workflows, integrating data, changing decision rights and building domain-specific AI systems—not simply adding AI assistants to existing processes. The leader of the future AI also changes what leadership means. When information, analysis and execution are increasingly available on demand, leaders cannot rely simply on having the answers. Their role shifts towards asking better questions, creating clarity, setting context, exercising judgment, aligning people and taking accountability. The future may be less about managing people or technology and more about orchestrating human and AI capabilities around outcomes. What happens to our human capabilities? Perhaps the most thought-provoking part of the conversation is Zafar's concern about cognitive over-dependence. We already outsource calculation to calculators and recall to search engines. AI takes this much further by allowing us to outsource drafting, analysis and even parts of reasoning. His question is not whether we should use AI—we clearly should—but: What mental capabilities do we deliberately continue to exercise ourselves? He argues that productive struggle—the process of wrestling with a problem, making mistakes and changing our minds—is part of how we develop critical thinking and intellectual independence. AI and the human relationship The conversation then moves from cognition to something even more fundamental: human relationships. AI is infinitely patient, always available and highly personalized. But what happens if frictionless AI relationships begin replacing the messy, challenging and sometimes uncomfortable relationships we have with real people? That raises a bigger question about human agency: are we eventually going to outsource not only memory and reasoning, but also judgment, empathy and emotional support? The responsibility that comes with intelligence Zafar also reflects on the societal implications of increasingly autonomous AI. AI has extraordinary dual-use potential—it can accelerate scientific discovery and healthcare, while also being used for surveillance, manipulation, cybercrime and misinformation. As AI becomes increasingly agentic, his message is clear: governance and human oversight need to advance alongside capability. He argues for a risk-based approach to AI governance: a tool rewriting an email should not face the same level of scrutiny as an AI making decisions about healthcare, employment, credit or public security. Ultimately, this is a conversation about much more than AI technology. It is about what happens to organizations when intelligence becomes abundant, what happens to leaders when answers become cheap, and what happens to humans when machines become increasingly capable. And perhaps the most important question we leave with is: As AI becomes better at thinking for us, what must we continue learning to think for ourselves?
My guest today is Paramu Kurumathur: IT Leader, Author. Here is the snippets of interesting stories around how AI has changed his life and how he’s impacting AI - From building AI in 1980 to using AI to code, write, research and imagine the future
My guest today is Yeshwant Satam, an ICT professional and consultant with over four decades of experience. In this conversation, he freely shares: His four-decade journey from soldering boards and PROM patching to firmware, software, IT services leadership, and manufacturing automation. Continual learning multiple programming languages including Pascal, Fortran, C, and Occam, then rapidly building capability through curiosity and commitment. Memorable moments including reverse engineering an IBM PC BIOS before realizing the boot floppy was still in a drawer in the US. The early days of Indian software exports: faxed requirements, shared processors, STPI floppy submissions, and the birth of professional delivery processes. That managers in technology must stay connected to the craft, even if they no longer code every day. Tips for great crisis leadership which includes staying calm, finding the real problem solvers, shielding the team from noise, and tracking technical progress closely. On AI in manufacturing, he distinguishes stable predictive AI in core operations from generative AI, which is currently safer at the fringes and in human-assisted decision making. Career advice for new entrants: study where hiring is moving, reskill seriously, and do not wait for a company to teach you everything. Career advice for mid-career professionals: get back in touch with technology, learn continuously, and show that you can move with new directions. Actionable takeaway: stay curious, stay close to the work, respect people who solve problems, and never let a title separate you from learning.
My guest today is Arun Subash Manickam. An Engineering leader at Google. In this conversation, Arun shares: his journey from Chennai schooling to computer science and a long career in networking, startups, acquisitions, and Google. transitions from individual contributor to manager, startup life, acquisitions, autonomy, AI, and career growth. learning continuously, adapting to change, and working hard with purpose as core career principles.