How is Quantitative System Development Different from Internet Development?
What’s the Difference Between Internet Algorithms and Quantitative Stock Selection?
......
Faced with various questions from candidates, let's make it all clear at once! To help everyone conduct a self-assessment when considering a career transition and plan for their financial future!
1. A Common Trait of Successful Transitions to Quant!
Quantitative investing, although also applying computer technology (including research and trading), has some differences compared to how the internet industry utilizes computers.
A simple analogy – Building a House
A real estate developer can build apartment buildings or villas.
Among the construction crews, some specialize in foundation work (using concrete), some in the main structure (using steel), some in external cladding (using panels), and others in interior finishing (using fixtures)......
If we view the process of quantitative investment's "R&D – Returns" as building a villa, and the internet's "Development – Product" as constructing a high-rise apartment building, then what kind of talent does quantitative finance want when recruiting from the internet?
Answer: General-purpose foundational talent – Low-level development!
No matter the type of building, the commonality is the need for a foundation, making foundation crews the most adaptable. Similarly, low-level developers from the internet, especially C++ developers, are more welcomed in the quantitative industry!
Conversely, if your previous experience in the internet industry involved application development and maintenance, or recall algorithms, recommendation algorithms, etc. – which is akin to working on the building's structure or interior finishing – your application scope in quantitative finance will be very limited!
After all, the "villa" quantitative finance wants to build is not the same as an ordinary building, and the required technical team differs.
2. What Kind of Low-Level Development Does Quant Like?
C++
Quantitative finance, especially high-frequency trading (HFT), has a strong preference for developers proficient in C++.
Performance Tuning Specialists
If your previous internet role involved I/O performance analysis and optimization, congratulations, your skills are a match!
Both quantitative trading systems and internet execution systems require continuous performance optimization to ensure smoother operation and reduce latency.
Therefore, if you have accumulated rich experience in improving I/O path performance, you will certainly be sought after by quantitative firms.
Compute Resource Schedulers
This is essentially also a system performance optimization role, focusing on compute resource scheduling for internet platforms to improve the efficiency of high-performance data sampling.
Compute resource allocation across different platforms (like K8S, ModelArts, etc.) is often dynamic. Talent in this area typically optimizes scheduling through techniques like shared information technology to enhance efficiency.
3. Optional Role Directions for Transitioning to Quant
When quantitative institutions look to hire from the internet industry, they generally focus on a few key areas:
- Trading System Development Engineer;
- Machine Learning System Engineer;
- Operations Engineer (DevOps).
Trading System Development Engineer
This is quite straightforward. No matter how excellent the strategy, it needs to be implemented within a computer trading system to achieve potential returns.
Typical Responsibilities:
- Responsible for the architecture design and construction of quantitative trading systems.
- Responsible for performance optimization across different compilers, operating systems, and network environments.
Typical Requirements:
- Some experience in low-latency system development.
- Proficient in data structures and algorithms, with solid C++/C programming skills.
- Mastery of advanced Linux kernel knowledge.
In short, the requirements for trading system developers definitely lean towards being more low-level!
Machine Learning System Engineer
This falls under the category of technology generalists. Using the earlier analogy, if you can draft blueprints for an apartment building, you can certainly draft blueprints for a villa!
Typical Responsibilities:
- Based on the internal compute cluster, build and optimize proprietary machine learning/deep learning training and inference frameworks to support efficient and accurate model training, deployment, and inference.
Typical Requirements:
- Excellent programming skills, proficient in C++ or Python, familiar with the underlying implementation of PyTorch/TensorFlow.
- Familiarity with CUDA, TensorRT, OpenVINO.
- Experience with AutoML aspects like NAS (Neural Architecture Search), HPO (Hyperparameter Optimization), etc.
Additionally, for MLE (Machine Learning Engineer) roles, candidates with experience at organizations like Google Brain, FAIR, MSR, DeepMind, OpenAI are highly sought after, representing top-tier industry expertise!
Operations Engineer (DevOps)
Compared to the previous two roles, the requirements for a DevOps engineer are broader, with a key emphasis on years of experience, particularly within tech companies.
This role splits into DevOps, MLOps, and more. Let's take MLOps as an example.
Typical Responsibilities:
- Responsible for the architectural design and core code development of quantitative computing platforms, such as one-stop machine learning platforms, MLOps, workflow systems similar to DolphinScheduler, etc.
- Development of internal productivity tools, optimizing the workflow for project delivery, exploring and defining work patterns for the cloud-native era.
Typical Requirements:
- Proficient in using JavaScript, HTML5, CSS3, etc.
- Understanding of MVC/MVVM patterns, familiarity with popular frameworks like React/Vue.
- Familiar with the use and configuration of frontend and backend build tools.
For general DevOps roles, you can probably fill in the details yourself. It should be quite straightforward for skilled technical professionals like you!
We'll stop here for now. Hope this helps our friends transitioning from the internet to quantitative finance!
Welcome to our Blog!
If you are interested in career opportunities, please send your CV to: recruiter@alphamaker.link
If you would like to share your story, please contact us on WeChat: AMQuantcareer