From SDE I to SDE III in 4 Years: What I Learned

In 2021, I joined Mamaearth as a frontend engineer right out of college. I knew React at a tutorial level — I could build a todo app but had never seen a production codebase. My first week, I opened the repository and felt genuinely overwhelmed. There were 400 components, a custom build pipeline I didn't understand, and patterns I'd never seen in any course.

Looking back, that feeling of overwhelm was the most productive state I've ever been in. It forced me to learn quickly, ask questions without ego, and develop systems for managing complexity that I still use today.

The most important skill I developed in my first year was learning to ship. Not writing perfect code — shipping working features that solve user problems. There's a gap between knowing a framework and building products with it, and the only way to bridge that gap is to build, ship, get feedback, and iterate.

At Mamaearth, I worked on the product listing pages, collection pages, and eventually led the performance optimization effort that reduced TBT by 60%. That performance work was my breakout project — it was technically challenging, had measurable impact, and gave me visibility across the engineering organization.

The lesson from year one: find a project that's important but unglamorous — something that senior engineers are too busy for but that has real impact. Performance optimization, test coverage, build tooling. These projects teach you the codebase deeply and demonstrate initiative.

I joined CARS24 as an SDE II and was promoted to SDE III within 18 months. The transition from IC to tech lead was the hardest part of my career. Suddenly, I was responsible not just for my code but for the team's technical decisions, architecture, and mentoring junior engineers.

The skill that mattered most at this stage was communication. Being able to explain technical trade-offs to product managers, write clear technical specs, and give actionable code review feedback. I spent as much time writing documents and having conversations as I did writing code.

At CARS24, I led the micro-frontends migration, built the design system, and established our performance budgets. Each project taught me something different: the migration taught me about architecture trade-offs, the design system taught me about developer experience as a product, and the performance work taught me about data-driven engineering.

Joining Flipkart was a step change in scale. The challenges aren't fundamentally different, but the blast radius of every decision is larger. A performance regression doesn't affect thousands of users — it affects millions. A bad architectural decision doesn't slow down one team — it slows down twenty.

The skill I'm developing now is systems thinking. Understanding how my team's work connects to the broader platform, identifying leverage points where a small effort creates disproportionate impact, and building tools and patterns that other teams can adopt.

At Flipkart, I work on the React Native applications for internal tools and the OTA update pipeline. The most rewarding work has been mentoring — helping junior engineers develop the same skills that accelerated my career. Teaching forces you to articulate things you've internalized, which deepens your own understanding.

First, your technical skills plateau faster than you think. After two years, the marginal return on learning another framework or library is much lower than the return on improving your communication, leadership, and system design skills. Technical depth is necessary but not sufficient for career growth.

Second, the job market rewards demonstrable impact more than technical sophistication. Nobody cares that you know five state management libraries. They care that you reduced page load time by 40% or improved team velocity by 30%. Quantify your impact and tell that story clearly.

Third, mentorship matters more than credentials. The senior engineers who invested time in reviewing my code, challenging my assumptions, and pushing me toward harder problems accelerated my growth more than any course or certification. Find those mentors and be that mentor for others.

Ship things. It doesn't matter if they're imperfect. The gap between knowing and doing is bridged only by doing. Build side projects, contribute to open source, volunteer for the project nobody wants. Every shipped project teaches you something a tutorial cannot.

Write about what you learn. Not for an audience — for yourself. Writing crystallizes your understanding and creates a body of work that compounds over time. These articles you're reading? They started as notes I wrote to myself after solving hard problems.

Be patient but not passive. Career growth isn't linear, and comparing your trajectory to others on Twitter is a recipe for misery. Focus on getting 1% better every day, and trust that the compound effect will be significant over time. Four years ago, I couldn't read a production codebase. Today I design systems that serve millions of users. The gap was bridged one commit at a time.