I recently completed an interview process for an SDE 3 position at J.P. Morgan Chase, which spanned three distinct rounds.

The first round was led by a senior associate, an SDE 3 themselves. It began with typical behavioral questions, after which I was presented with a code snippet. My task was to review this code, identify potential issues, and suggest improvements. This round concluded with several questions focused on core Java concepts.

My second round was with a Principal Engineer, an Executive Director. The main challenge here was to design a Balance API service capable of ingesting events from other APIs and accurately maintaining user balances. We delved deep into various design choices and considerations. I felt this interview went particularly well, largely due to the interviewer's expertise and helpful demeanor.

The third and final round was conducted by a VP of Engineering. This round also started with behavioral questions, followed by discussions about my past projects. Interestingly, the interviewer referenced the code snippet from my first round, posing further questions specifically about concurrency and thread safety in that context. Like the first round, it wrapped up with more core Java questions.

Later, the recruiter informed me that the offer was for an SDE2 position, a downgrade from the SDE3 role I had interviewed for. I decided to reject the offer, as I am currently a Senior Software Engineer and did not wish for a career downgrade on my CV, especially since the compensation offered was very similar to my current package.

I interviewed for a Senior Associate Machine Learning position at JP Morgan Chase. With over 7 years of experience primarily in AI/ML, focusing on NLP, OCR, AWS Sagemaker, Azure Machine Learning, and some basic computer vision, I was eager for this role. The interview process itself was a bit challenging to schedule, with the call being rescheduled 5-6 times before it finally took place. When the 45-minute call began, the interviewer joined about 5 minutes late. Approximately 10-15 minutes into the session, a coding question was presented. The interviewer directly shared a LeetCode link to the problem "Subarray Sum Equals K." I struggled with the problem; initially, I failed to consider zero values, which I then fixed. I received a hint from the interviewer to use a hash map, but unfortunately, I then made a mistake by considering an O(n) array memory solution. Afterward, I failed to account for negative values. By the time I was attempting to fix these issues, the allocated time for the problem was over. The interview did not cover any topics related to AI/ML, NLP, GenAI, RAG, or Multi-agents, focusing instead on Python DS & Algo. I was subsequently rejected.

All happened in October.

Compensation : 40 LPA Fixed

I applied to on of the positions on their Job portal.

# OA
Q1: Minimum operations to make the 2nd half of the string an anagram of the first half. swapping 1 character with another is 1 operation.
Q2: Given queries as [L,R], flip the sign of all the elements of array from L to R. Return the array at the end.

After this got call for cohort hiring and all 3 round will happen back to back.

# Coding+Code review round
Code Review a piece of code: There were a few APIs and you need to review it - check variable names, exception handling, using enums, reusing common code, synchronization and locking etc
Coding : Problem (telephone circular dial)

You have a circular dial with digits 0,1,2,...,9 arranged in that order around the circle. The dial’s pointer starts at 0. You are given a string s of digits (for example "8053") that you must dial in order.

You may rotate the pointer clockwise or counterclockwise. Moving the pointer by one position (to an adjacent digit) costs 1.

Optionally, pressing a digit might cost 1 (if the problem counts presses). I'll show both variants: rotation cost only and rotation + press cost.

Return the minimum total cost to input the whole string.

# HLD Round
Asked to design a Rate Limiter
Got into an agrument in this, during requirement i had said I will implement the sliding window counter method (which will be an approximation) as I dont think the rate limiting needs to be completele accurate.
But during designing that, she started saying read the question again (which didnt mention anyting strict as requirement) I said the same no such thing is mentioned and I explained already my assumptions during requirement, Anyway I said if we want it to be strict we will need a sliding window log method which will use memory.

# Behavior Round
Several questions regarding my project and how i handled this and that etc.

# Verdict: Rejected (probably)
Never got an update after the interviews. and the HR never picked my call again. HR never shared his email and communication happened over jpmc system generated mails.

I have around 2.5 years of industry experience, and I recently appeared for the Software Engineer 2 (Associate) position at JPMorgan Chase, Bangalore. The role was for a Java/Python Developer, but all the assessments and interviews were conducted strictly in Java — no other language option was provided.

The entire interview journey was quite well-structured and covered multiple aspects of software engineering — from coding and system design to code review and behavioral discussions. After the technical rounds, there was also a team matching discussion post-offer to finalize the placement within a team.

Here’s a detailed breakdown of my experience

Online Assessment (OA)

There were 2 medium-level coding questions.

Both were quite similar to the “Maximum Number of Overlapping Intervals” problem — the logic was essentially the same, though the wording and context were different.

Round 1 – DSA + PR Review

This round was conducted on HackerRank CodePair.

It began with a Pull Request (PR) review task.

I was asked to review a piece of code, identify issues, and suggest improvements.

My feedback included:

• Fixing compile-time errors

Improving variable naming conventions

Applying proper data-hiding concepts

Writing cleaner, modular functions

I added my comments directly in the code.

After that, the interviewer asked me to run the code — there was an import issue I couldn’t fully resolve, but since the rest of my feedback was strong, we moved on to the DSA question.

```
You are given k circular dials (each containing characters ‘A’–‘Z’) and a target string s.
Initially, all dials point to ‘A’.

In one move, you may rotate any one dial clockwise or counter-clockwise by one step.

You can type a letter only if at least one dial currently points to that letter.

Find the minimum total number of moves required to type the string s sequentially.

The discussion revolved around optimizing both time and space complexity while maintaining clean, readable code.

Round 2 – System Design

This was a high-level design discussion on building a Rate Limiter.

We covered:

• Functional and non-functional requirements
• API design and data flow
• Database schema
• Cost estimation and scalability
• Full HLD diagram and component interaction

We also discussed use of caching (why and what type: read-through vs write-back), API call patterns and performance considerations and horizontal scaling strategies. Overall, it was an in-depth and engaging discussion.

Round 3 – Behavioral

This was a standard behavioral interview with questions like:

• What challenges did you face in your recent projects, and how did you overcome them?
• Describe an innovative solution you implemented.
• How do you handle conflicts with teammates or managers?

The interviewer mainly assessed communication, ownership, and problem-solving mindset.

Round 4 – Team Matching (Post- Offer Discussion)

After receiving the offer, I had a Team Matching round where I spoke with a potential hiring manager.

The discussion was centered around:

• My previous projects and technical contributions
• Why I wanted to join JPMC
• Reasons for moving on from my previous organization
• My interests and preferred tech stack

It was more of a mutual fitment discussion than an evaluation round.

Verdict: Selected 🎉

I applied for various SWE 2/3 roles on the JPMorgan Chase careers website and received an Online Assessment (OA) for one of them.

Round 1: Online Assessment (HackerRank)

The OA consisted of two questions, which I recall being LeetCode Easy to Medium difficulty. I don't remember the exact problems.

After clearing the OA, I received a call from the recruiter to schedule my Super Day, which meant all three virtual interview rounds would happen back-to-back on a single day.

Virtual Interviews (Super Day)

Round 1 (Hiring Manager Round)

This round was conducted by a Director at JP, who had approximately 17 years of experience. We discussed how I would approach different challenges within a team environment and how I would manage those with clients, given my client-facing experience mentioned on my resume. The interviewer was quite relaxed, and I felt he was assessing my spontaneity rather than just my prepared answers.

Round 2 (Data Structures & Algorithms)

In this round, I was first given a Pull Request to analyze. I spent about 15 minutes reviewing the code and pointed out issues such as hardcoding values that should be in environment files, the plain use of passwords instead of hashing, and potential design patterns that could be followed. Following this, I was given a medium-level DSA question on arrays, where the task was to sort a string based on the order of occurrences of its characters. I then had to explain my choice of data structures, along with the time and space complexity of my solution.

Round 3 (High-Level Design)

This round primarily focused on system design. The interviewer, a senior colleague who seemed busy, asked me to design an end-to-end payment processing system. I explained everything in detail, covering requirements, database schema, estimations, the high-level design, and potential further optimizations. This was all done on a CodePair Link.

I recently completed my first-round interview for a Java Developer position at JP Morgan Chase in Bangalore. The interview covered a range of topics, starting with several behavioral questions focused on my career motivations, problem-solving skills, and teamwork experiences. Following this, we delved into Java-specific technical questions, primarily revolving around concepts like synchronization, locking mechanisms, and the characteristics of ConcurrentHashMap and exceptions. The final segment of the interview involved a code review. I was presented with a class designed to manage account balances in a map, with methods for updating and retrieving balances. During the review, I suggested using ConcurrentHashMap for thread safety, which was acknowledged as correct, but I struggled to elaborate on its internal implementation details. My responses to the 'why JPMC' question and my recollection of the job description were not strong, as my current project work is primarily in C++. Given these factors, I don't expect to move forward in the interview process.

I was initially approached by a recruiter on LinkedIn for a Software Engineer 2 position at JPMorgan Chase. With 3 years and 2 months of experience in the Fintech domain and a TC of 20 LPA, I decided to proceed with the interview process.

Round 1

This round was taken by a VP and started with several questions from my resume, primarily focused on my experience. After that, I was asked to open Hackerrank and solve an OOPS-related question. This led to a detailed discussion with follow-up questions, mainly regarding HashMap internals and the proper overriding of equal() and hashCode() methods. The interviewer also asked a semi-LeetCode style question involving array manipulation, though I don't recall the exact problem. Towards the end, we discussed general questions about Kafka and REST APIs, both of which were mentioned on my resume.

Round 2

This round was also conducted by a VP. After brief introductions, I was asked to directly open the Hackerrank link. I was then given two LeetCode medium-level questions, both related to array manipulation. Unfortunately, I don't remember the exact problems.

Round 3

The third round was again with a VP, and it was quite intense. I was thoroughly grilled on Java internals, specifically focusing on multithreading concepts. This included questions about the volatile keyword, how it works internally, the differences between optimistic and pessimistic locking, and the internal mechanisms of the synchronized keyword. Following that, the discussion shifted to Kafka internals – how Brokers function, how they maintain topics, strategies to increase Kafka's throughput, and the concept of Dead-Letter Queues (DLQs). Finally, the interviewer delved into Webflux internals, which I had mentioned on my resume, leading to some questions about Virtual Threads.

I tried my absolute best in this round, but it was personally very challenging for me. I ultimately washed out from this round and did not receive an offer.