‘A strawman proposal is a proposed solution to solve a problem’: One-on-one with Silicon Valley veteran Pawan Kumar
Pawan Kumar was a long-time upper management figure in engineering expertise throughout the software industry for 20 years; he is currently settled in Silicon Valley. He received a sturdy background in technology, graduating from the Birla Technical Training Institute in Pilani, India; his thorough experience ranges from architecture design, engineering management to core engineering and the cloud. He has put effort into different industrial domains like airlines, banks, insurance, healthcare, and advertising and has seen multiple software practices applied there.
Q 1: What motivated you to pursue a career in software engineering and eventually engineering leadership?
A: My interest in software engineering began because I liked solving complex problems with technology. The field was interesting as it was continually changing and moving. It was an important transforming force for business and the user–two fields that I naturally gravitated toward once I started down my occupational path. The evolution of my career let me rotate into supervising roles for larger and more complex systems, in which I was successful initially. Hence the natural path not long before I completely committed to mentoring teams and architecting scalable solutions. Leading high-octane engineering teams in a technology company allows you to multiply the impact by empowering others while still staying technically interconnected to the innovation and trendiest technology.
Q 2: How do you approach scaling high-performance systems, and what key factors do you consider?
A: I always approach this with an unbroken focus, understanding not only business requirements but also why technical constraints are the business. I always focus first on making well-distributed patterns, which have all but considerably relieved load and are highly recoverable. The main factors are distributed data strategies and intelligent read-write data workflows, coupled with caching mechanisms, and smartly configuring an event-based asynchronous style of processing to guarantee that you build systems where only single points of failure are designed with redundancy and fail-over capabilities. In my career, I have worked with technologies like Cassandra, Kafka, and various cloud platforms which can handle thousands of operations per minute while keeping high availability in multiple data centers.
Q 3: Can you describe a challenging project you managed and how you overcame obstacles?
A: The most taxing project my hands had to manage was revamping a legacy system into a modern microservices architecture. We stumbled across massive hindrances during the phase of data migration, backward compatibility, and zero downtime during the changeover. To overcome this long list of problems, I decided to adopt a three-fold migration, so parallel runs of both systems were processed, packed full with testing protocols, and rollback plans. My set-up included a smooth stream of information between all stakeholders and I was able to crush any daily sync meetings to solve roadblocks. By breaking up the migration into small work packages and very carefully orchestrating steps, we were able to totally and successfully modernize the platform while achieving way better performance improvements and minimizing most of the disturbance.
Q 4: What role does mentorship play in your leadership approach?
A: Mentorship plays an integral role in my leadership philosophy. Team members should be given the opportunity to learn and grow, both professionally and technically. I hold regular one-on-one sessions with team members so as to grasp their aspirations and challenges to aid them with the help of some sound advice. I also organize opportunities to work on great assignments that stretch an engineer’s abilities, with providing only some amount of the necessary support. Investing in training my team has meant huge achievements in innovative approaches, team attitude, and retaining gifted expertise.
Q 5: How do you integrate emerging technologies into the legacy system?
A: The solution to incorporating emergent technology lays in a fine balance between innovation and stability. I begin by working on a proof of concept as a “test bed” for understanding the considerations of a new technology within our own context. Moreover, when the new technology is integrated, I prefer to proceed incrementally often from not-too-risky components of it: for instance, if we are moving from some different databases to NoSQL databases like Cassandra or Elasticsearch, we introduce them incrementally for selective functionality. I have the capability to transition monitoring and/or fallbacks during the course of this. I have successfully introduced the likes of cloud services, containerization, and distributed data stores through this diligent, technology- smart, and business-smart approach while still leaving default enterprise systems as they were.
Q 6: Would you give your favorite tools or frameworks to bring in robust software systems?
A: There is a selection of tools that I return to frequently in addressing robust solutions. I have been using Java as the backend programming language paired with Spring in the arsenal for long years now owing to its maturity and powerful framework features. In terms of data persistence, I have used a historical combination of the traditional relational database with NoSQL systems such as Cassandra, Elasticsearch, or other distributed caching technologies for specific use cases. Event-driven architecture and messaging, which has been the lifeline, I made use of Kafka. In my maximum experience, I have worked on cloud infrastructures many times, having also been present in the use of their services. I usually use the ELK stack with comprehensive logging for monitoring and observability. Apart from getting overexcited about development, I shall advise persons to choose the most appropriate tool for the job instead of cramming in a wrong-fit technology.
Q 7: How do you weigh any technical debt diminishing with building any new feature?
A: The challenge is continuous and requires strategic planning; hence, this topic is alive. I approach a balance between technical debt reduction and new feature development by building it into a normal development cycle rather than solely as a separate activity. Taking roughly 20% of the capacity of any sprint solely for clearing technical debt in which we primarily focus on areas that give us the highest return in terms of maintenance and further development speed, technical debt visibility is maintained by holding a separate backlog and establishing metrics to evaluate its impact. Certain technical standards established for the quality of new features helps prevent an increase in technical debt as they are consistently and carefully followed. Balancing the two ensures achieving business value while incrementally strengthening the technical foundation.
Q 8: What words of counsel would you render to someone aiming at growing in software engineering with the ultimate aim of becoming a lead?
A: Well, I would tell him or her to be skilled in what they are doing. Get your head down to the serious study of the real, engineering basics, and secondly, acquire some level of Mastery in at least one ought field. Learn to communicate with the business – share the intrinsic capability of explaining complex entities simply, listen well, and follow requirements quickly. And also, if possible, try to be ahead of the smaller projects by taking up little projects here and there. Try at the same time, whilst being led by mentors for your growth, to mentor others. Technology shifts so promptly that staying curious, and adapting to newer skillsets, are virtues that engineer prospective leadership. Remember that leadership is, above all else, technical facilitation, teamwork building, and fostering individual growth itself.
Q 9: What is your method of remaining in touch with industry trends and new technologies?
A: The answer to this question is pretty straightforward. It requires few ingredients: an individual character, some hard work, curiosity, and the thirst for achieving something, and, most important, a true passion for engineering. It is all this that keeps your step just a pace ahead ahead of others. Every week looks set for tech blog updates, magazines, and research papers. I am heavily involved in a myriad of online communities and keep up with influential thought leaders in software engineering. The awakening of my learning comes from countless conferences and meetups, webinars, and even casual chats with peers, where I seek solutions relevant to current problems. Since the long- serving motto “Learning by Doing” proved the best of many practices in my life, I can say I have remained in the game all through. The knowledge from teaching and mentoring also strengthens my understanding. Costume-learning networks and coherent continuous-learning strategies have helped me remain ahead in industry trends throughout my career.
Q 10: What are your views and opinions on building resilient distributed systems?
Answer: When building resilient distributed systems, it is essential to appreciate that failure is a reality, rather than an exception. My focus is on designing systems that can degrade gracefully in the face of component failure, instead of failing completely. This means implementing patterns like circuit breakers, bulkheads, timeouts, and retries across service boundaries. Consistency of data in distributed environments is particularly tricky, and I choose, based on a business assessment, eventually consistence or strong consistency. The comprehensive monitoring, observability, and automated recovery mechanisms necessary for the system’s health can be prepared. The chaos engineering practice becomes quite valuable at this juncture because it helps in identifying hidden weaknesses at an early stage. Throughout my career, I have learned that resilience is as much a question of team processes and culture as it is of technology – teams need to be quick to respond to incidents and consistently learn from their failures.
About Pawan Kumar
Pawan Kumar brings with him more than 20 years as an engineering leader in the software sector, having graduated with a Bachelor’s degree in Information Technology Engineering from Birla Institute of Technology and Science in Pilani, India. Pawan built a fine career in building enterprise scale applications in a multitude of sectors.
His expertise includes Java/J2EE technologies, distributed systems, cloud architecture, and engineering management. Pawan has effectively led teams through the task of designing and implementing high-performance scalable systems that cater to thousands of transactions per minute. He is also very much committed to mentoring software engineers and building high-performing teams that deliver exceptional results.
News