Physics Wallah Staff Backend Engineer Experience

I have 7+ years of backend experience, most of my preparation for the Physics Wallah process was around system design and the kind of engineering decisions I had taken in my current and previous projects.

Round-by-Round Experience

Initial Screening: Hiring Manager

My first conversation was with the hiring manager, spent most of the time discussing my current project, responsibilities and the kind of systems I was working on. The HM also explained the role, the team and what they were looking for at the Staff Engineer level.

It was more of an introductory technical discussion before moving into the main interviews.

Round 1: DSA

The coding round had a problem around finding the Longest Increasing Subsequence in an integer array.

I first discussed O(n²) dynamic programming solution, where dp[i] represents the longest increasing subsequence ending at index i.

The interviewer then asked whether I could improve it.

I moved to the O(n log n) approach using binary search, maintaining a structure where each position represents the smallest possible ending value for an increasing subsequence of that length.

There were follow-ups around edge cases, duplicates and the complexity of both approaches.

Round 2: Experimentation and Feature Flag System

This was my first HLD round. The requirement was to design an experimentation/feature-flag platform capable of handling around 2 million requests per minute. Started with how an internal service would determine which experiment or feature variant should be shown to a particular user.

One important part of the discussion was avoiding a network call to the experimentation service for every request.

I proposed providing a client library/SDK to internal services so that most evaluations could happen locally.

For assigning users consistently to variants, we discussed consistent hashing, so the same user would continue seeing the same variant.

Another interesting discussion was around configuration updates. When an experiment or feature flag changed, those updates needed to reach the client libraries quickly. I discussed using a CDC-based propagation mechanism rather than having every client continuously query the database.

From there, we went deeper into caching, failure scenarios, consistency and how the design would behave as traffic increased.

Round 3: E-Book Platform at Large Scale

My next design problem was an e-book platform supporting around 6 million concurrent readers during a major launch.

I had to support browsing, purchasing, free books, previews and protected access to paid content. I discussed the data model, APIs, authorization, preview access, CDN usage and caching strategy.

A major part of my discussion was deciding what I could cache aggressively while still ensuring that users couldn't access protected book content without authorization.

Round 4: Senior VP Discussion

My final conversation was with a Senior VP.

Instead of receiving a fixed design question, I picked one of the most technically challenging projects from my experience and walked through its architecture and key decisions.

I was then given changing scenarios such as:

What would I do if the traffic increased by 10x, and what part of my system would fail first?

I explained how I would identify bottlenecks across databases, networking and caching before deciding where additional infrastructure was actually required.

Interview Result

After a week of the process , I received a mail , selected for the Staff Backend Engineer role.

Tips

While DSA is definitely important, I would suggest spending more time on system design, especially for senior-level roles.

In my experience, the interviewers were not satisfied with just a working architecture. They kept changing the scale and constraints to see how I would adapt my design.

So, while preparing, make sure you can explain why your design works, where it might fail as the system grows, and what changes you would make to handle the next level of scale.

Being able to justify your decisions and discuss trade-offs is just as important as designing the system itself.