Self evaluation

Capitalab

Position: Quant analyst
Time: June 2022
Reached Stage: Final Round
Process: Screen chat, Python take home (poker hands), Technical Screening (1 math + 1 sde chats), Final rounds(met 4 people)
Mistakes: Last interview with the CEO (Gavin) could not solve the math question integrate logx... Forgot integration by parts...
Reject reason: stupid math question + not enough finance background. They asked lots of option pricing stuff. Volatility trading. Black Scholes is wrong.. Implied volatility
Their business: portfolio optimization + compression, greek modelling. very niched business. Rates compression, derivatives, small company. Swaptionizer, swap on options. Capital Lab is under BGC. optimization problem, loss functions, constraint programming high paced, entrepreneurial environment,regulations → need for optimization, no regulation = no money for CapitaLab

World Quant

Position: Quant developer - write internal services and automating tools for quants and portfolio managers
Time: June 2022
Reached Stage: Screening
Process: Screening, Hackerrank, Technical screening, 3 x 45 min tech screenings, Final interview with general manager (James)
Structure: Quant Researchers + Portfolio managers. Quant Devs support both. Small 6month projects. Simulator written in C++.
Tech Stack: Python, C++, Linux, some AWS.
Reject reason: Not enough experience

Trailstone Group

Position: Quant developer / researcher. Role is to champion best practices, code a lot, scalable data tools. Good testing, modularity, maintainability, flexible code, no over engineering. Understand whole business, interact with Front Office. Main role is in Berlin, there will be all people of the company mainly. Automate management based on scalable technology
Time: June 2022
Reached Stage: Final Round
Process: Screen chat, Hackerrank (algo + pandas), Technical (algorithms with Michael), Final rounds(met whole team ) They did not make me quant dev offer and wanted to interview me for quant research (said bullish on making an offer)
Mistakes: On final round they asked what problems would you expect to have when developing a trading strategy and implementing it in practice. Quant devs take care of that. Data problem, services, notifications the whole pipeline. I did not know this stuff.
Reject reason: For QR they asked me about linear regression and how points affect the line. Larger Residuals have larger i mpact (squared error)... Also the x-axis position of he point matter as well. The farer you are from the mean the larger impact you have on the regression line. Question about optimization problems: Linear regression with beta<5 (yes) is that a convex problem. abs(beta) < 5 is that a convex problem (no - you might be on the wrong side of the parabolla). Dual problems, how to know you are on global minimum - maximize your dual problem = minimizing your loss
Questions: Agile, SQL, Python, Linear regression, Optimization problems
Their business: build renewable energy platform, commodity hedge fund

Hanson Applied Sciences - Smarkets

Position: Quant developer
Time: June 2022
Reached Stage: Offer
Process: Screen chat, Hackerrank (algo + math), Technical Screening (1 algo + 1 ML chats), Final round (went to office, HR chat + chats with Team + great lunch!)
Mistakes: In negotiation stage, when they made me an offer I should have asked right away if the offer is negotiable. Always do that, so that you can set expectations for them that you want more. Got in small car/bike accident on the way to the interview. Don't put your headphones when driving a bike + be concentrated on the driving when doing it.
Their business: Market making on sports betting exchanges Hanson Applied Sciences – is a from Liquidity provider turned to Market Maker Part of Smarkets 140 people/ 70 technical staff Online betting Quant dev build production code (close to engineering role) team of 3 quant devs graduate, hedge fund guy, sport betting guy Tech stack Python(quants), Rust (engineers), SBK Sports book keeper is a mobile app goes through their other product which is the exchange Quant dev is not customer facing – work with engineers
Pricing engine, lots of microservices, service for post trading analysis, no ML yet, something to be created
main rival is Betfair.
Hanson has two jobs make money, make sure all products are prices (provide liquidity). Lowest margins in the business, rely on large volumes. Technical + fundamental analysis.
People met: Sean, Qiang, Anthony, Flavio, Sarah, Simon, Ajay

Google

Position: Product Analyst Merchant
Time: June 2022
Reached Stage: First screening
Reason for reject: Talked with recruiter, Had great interview with product analyst, they have not answered me yet, I think there is hiring freeze (job ad is gone)
Role: product analyst gives direction of general business. No need to be super statistically rigorous. Write lots of SQL. Passionate about data. Talked with Haifend Wu in the interview. Shape product in strategic level. Do analysis/metrics, create dashboards/segmentation fof google's users.

Yelp

Position: Data Scientist
Time: June 2022
Reached Stage: Final Round
Process: Screen chat, Hackerrank (SQL + Python), Technical Screening (SQL + DataScience (A/B testing)), final interviews (Product interview, 1 presentation, 2 Data Science interviews, 1 HR/Leadership chat)
Mistakes: Look for keywords, retention, 2 things to measure in these places user experience and money. Key metrics (HEART) Presentation talked about ONE project, but could not say whats the final business impact. Overall the 2 Data Science interviews went 50/50 the rest were excellent. Don't interrupt your interviewer! Felt a bit off after the negotiation process. They made me good deal and I was hoping to bump a bit more from it, but god bump just 5k. Final offer 120 gbp ~ 160 usd which is good. However, I felt I could do better and get more money.
Role/Company: Tech Stack: Python, some R, Tableau, AWS for compute and storage, SQL. Role is to work with Multi-local business (they would be your clients). Unique dataset with lots of interactions

Marshall Wace

Position: Data Engineer
Time: June 2022
Reached Stage: Pre-Final Round
Process: Screen chat, Hackerrank (1 easy + 2 mediums), Technical Screening (1 algo (leetcode hard with 20% acceptance)

1 data chats), Final rounds

Mistakes: I could not quickly understand the hints during the data interview. It was simple SQL stuff with data table schema. Algo was leetcode hard and I did it. Communication barrier mistake as I could not get what they meant quickly. Those guys were data engineers but the feel was like they were traders/pretty sharp, young and quick.
Role/Company: small/productive teams structure, no BA-s, no PM-s, entrepreneurial, OWN-ership, make your own decisions, fast-paced environment, flat structure, serve quants mostly

Atomico

Position: Data Scientist
Time: June 2022
Reached Stage: Second technical round
Process: Screen chat, First technical (easy algos), Second technical, Take home, Panel/presentation, CEO chat
Mistakes: Weird interview in Jupyter notebooks, working with datasets. Could not click with interviewer and could not grasp quickly his ideas. Gave me data which was generated from sin((x-const)/const) + N(0,1) and I had to determine that. Did some visualization stuff, computed correlations, but did not have a structured way to get the final result so I was slow. I did not enjoy the experience either. Made mistake not to listen Arlo as I should have. Do not interrupt!

Shopify

Position: Senior SDE
Time: June 2022
Reached Stage: Schedule for Final Round
Process: Technical (algo) super simple hashmap, Life Story (1h chat, went good), Final Stage
Reject reason: Hiring freeze

Point72

Position: Data Scientist
Time: June 2022
Reached Stage: Take home task
Process: Hackerrank, Take home DS task
Mistakes: Did not heard from them after sending the take home task. It was not really good.

Akuna

Position: Quant Developer
Time: June 2022
Reached Stage: Final Stage
Process: Hackerrank, Technical(algo), Final rounds (presentation algo + math with QRs)
Mistakes: On the Hackerrank I was not concentrated and missed some parts even due to time pressure but still passed wtf

HRT

Position: Algo engineer
Time: June 2022
Reached Stage: Second stage
Process: Hackerrank, Technical (asked about a sde project, multiprocessing vs multithreading, + hard dp)
Mistakes: Could not solve DP

Given an array of size n and and integer k. We partition the array in k subarrays (some can be empty). Compute each sum of the elements in each subarray s1, s2, ... sk. Take the max. Among all partitionings minimize the max value.

Multiway partitioning problem

Leetcode

NP-hard

DE Shaw

Position software engineer ( the guy said could be 50/50 computer science vs maths)
Time June 2022
Reached Stage: Second
Process: Hackerrank, Technical in coder pad (gave calculator question + implement a queue with an array, table resizing...)
Mistakes: so far none?