New opportunity

Junior Quantitative Researcher | Crypto Delta One

BlockTech · Singapore

About this role

BlockTech is a fast-paced algorithmic trading firm operating at the frontier of global cryptocurrency derivatives and spot markets. We are seeking a Junior Quantitative Researcher to join our trading floor and contribute to building smarter models, sharper signals, and adaptive systems. Role Purpose: You will own research ideas end-to-end, from hypothesis and dataset construction through feature engineering, model training, and backtesting to live deployment and iterative improvement. You will work closely with Quantitative Traders and Analysts to drive pricing and trading strategies. Responsibilities: Collaborate with traders to translate research insights into systematic trading strategies. Design, develop, and deploy models for price prediction, signal generation, execution, and anomaly detection using state-of-the-art AI and ML techniques. Build robust trading, backtesting, and research infrastructure. Monitor live model performance, diagnose decay, and iterate on production systems. Required Qualifications: Strong academic background in a quantitative discipline (Mathematics, Physics, Statistics, Computer Science, Econometrics, ML/AI, or similar), typically a PhD or MSc with strong research experience. Fluency in Python and the modern ML stack (PyTorch, TensorFlow, scikit-learn, NumPy, pandas). Deep understanding of overfitting, generalization, validation design, and feature engineering. Solid software engineering instincts for writing readable, testable, and production-ready code. Excellent communication skills in English. Preferred Qualifications: Hands-on experience building and deploying ML models, particularly for time-series, forecasting, or anomaly detection. Familiarity with crypto markets, derivatives pricing, or high-frequency trading data. Benefits: Competitive base salary with an attractive performance-based bonus plan. Opportunity to buy into the trading fund for outstanding performance. Extensive in-house training and annual learning & development budget. Daily breakfast, lunch, snacks, and drinks. Gym membership reimbursement and sports event support. All-in relocation package for international candidates. The role is based in our Singapore CBD office.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

PythonPyTorchTensorFlowscikit-learnNumPypandasMachine LearningTime-series analysisForecastingAnomaly detectionAlgorithmic tradingQuantitative researchFeature engineeringModel backtesting

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

End-to-End Model Lifecycle Ownership
The role requires managing the entire research pipeline, from initial hypothesis and dataset construction to model training, backtesting, and live deployment. Candidates must demonstrate the ability to maintain accountability for a model's performance after it is live, including monitoring for decay and performing iterative improvements.
Robust Machine Learning and Statistical Rigor
Beyond simply applying models, the researcher must possess a deep, intuitive understanding of overfitting, generalization, and validation design. The firm prioritizes candidates who can explain the underlying mechanics of why a model works, ensuring that signals are based on sound statistical principles rather than noise.
Production-Ready Software Engineering
The researcher must write clean, testable, and readable code that integrates seamlessly into the firm's trading infrastructure. This requires balancing research-oriented experimentation with the software engineering discipline necessary for high-frequency, production-grade trading systems.
Cross-Functional Collaboration
Success depends on the ability to translate complex research insights into actionable trading strategies while working closely with traders and engineers. Clear communication is essential for explaining model behavior and strategy logic to a mixed audience of technical and non-technical stakeholders.

Likely challenges to think through

Managing Model Decay in High-Frequency Environments
Cryptocurrency markets are fast-moving and prone to rapid regime shifts. A significant challenge is distinguishing between temporary market noise and genuine model decay. Prepare by practicing how to diagnose performance drops and how to design adaptive systems that remain robust under changing market conditions.
Balancing Research Agility with Infrastructure Constraints
There is a inherent tension between the need for rapid experimentation and the requirement for stable, production-ready code. Prepare by demonstrating experience in building modular research pipelines that allow for quick iteration without compromising the safety or reliability of the live trading environment.
Navigating Complex Market Microstructure
The role involves working with crypto derivatives and spot data, which can be noisy and structurally unique. Prepare by researching common challenges in crypto microstructure, such as order book dynamics and latency, and be ready to discuss how you would design features that capture meaningful signals while avoiding overfitting to artifacts in the data.

Practice your answers

Technical practice

Handling Overfitting in Financial Time-Series

What to coverExplain your approach to preventing overfitting when training models on noisy financial data. Discuss techniques like cross-validation for time-series, regularization, and the importance of feature selection to ensure model generalization.

Feature Engineering for Market Microstructure

What to coverDescribe how you would construct features from high-frequency order book data to predict short-term price movements. Focus on how you capture market dynamics like liquidity, order flow imbalance, or volatility.

Model Decay and Production Monitoring

What to coverHow do you detect and diagnose performance decay in a live trading model? Suggest metrics or statistical tests you would use to identify when a model's predictive power has degraded due to changing market regimes.

Backtesting Infrastructure Design

What to coverWhat are the critical components of a robust backtesting system to avoid look-ahead bias and ensure realistic execution simulation? Discuss how you handle transaction costs, slippage, and latency in your simulations.

Machine Learning for Anomaly Detection

What to coverPropose an ML approach for detecting anomalies in crypto market data. Explain the trade-offs between unsupervised methods like isolation forests versus supervised approaches if labeled data were available.

Python Performance Optimization

What to coverGiven the need for low-latency execution, how do you optimize Python code for heavy numerical computation? Mention tools like NumPy vectorization, Numba, or profiling techniques to identify bottlenecks in your research pipeline.

Validation Design for Trading Models

What to coverHow do you design a validation strategy that respects the temporal nature of financial data? Explain why standard k-fold cross-validation is often inappropriate and what alternatives you would use.

Deep Learning vs. Traditional Statistical Models

What to coverCompare the use of deep learning (e.g., LSTMs or Transformers) versus traditional statistical models (e.g., ARIMA or GARCH) for price forecasting. Discuss the trade-offs regarding interpretability, data requirements, and computational complexity.

Behavioural practice

Collaborating with Traders

What to coverDescribe a situation where you had to translate a complex technical research insight into a practical strategy for a non-technical stakeholder or trader. Focus on how you communicated the trade-offs and risks involved.

Owning Research End-to-End

What to coverProvide an example of a research project you owned from hypothesis to deployment. Highlight how you managed the transition from initial exploration to production-ready code.

Handling Failure in Research

What to coverTell me about a time when a model you developed failed to perform as expected in a live or simulated environment. How did you diagnose the issue and what did you learn from the experience?

Prioritizing Research Tasks

What to coverIn a fast-paced trading environment, how do you decide which research ideas to pursue and which to abandon? Give an example of how you balanced the need for quick results with the requirement for rigorous testing.

Sharpen your knowledge

Topics to prepare

Model Validation and Overfitting
The role requires a deep understanding of why models work, specifically regarding generalization and validation design. Exercise: Review your past research projects and document how you specifically prevented data leakage and overfitting in time-series data, ensuring your validation strategy mimics real-world production constraints.
Market Microstructure and Financial Data
The firm emphasizes crypto derivatives and spot markets. Exercise: Study the mechanics of order books, limit order books, and trade execution. Research how high-frequency data differs from standard time-series data and prepare to discuss how you would handle non-stationarity in financial signals.

Skills to revise

Production-Ready Python Development
The role requires writing code that is readable, testable, and deployable. Exercise: Refactor a previous research script into a modular, object-oriented Python package. Include unit tests for data processing pipelines and ensure your code follows PEP 8 standards to demonstrate software engineering instincts.
Time-Series Feature Engineering
You will be responsible for end-to-end model development. Exercise: Take a raw financial dataset and perform feature engineering to create predictive signals. Focus on creating stationary features and lag-based indicators, then evaluate their predictive power using scikit-learn or PyTorch.
Model Monitoring and Decay Diagnosis
The role involves owning models in production and diagnosing performance decay. Exercise: Create a simple dashboard or logging script that tracks model performance metrics over time. Simulate a scenario where model performance degrades due to market regime shifts and document your process for identifying the root cause.

Your preparation checklist

Prepare a research project walkthrough
Select a past research project involving time-series forecasting or anomaly detection. Be prepared to explain your hypothesis, dataset construction, feature engineering choices, and how you specifically addressed overfitting and generalization.
Demonstrate production-ready coding standards
Review your Python code to ensure it follows clean coding practices. Prepare examples that demonstrate modularity, unit testing, and efficient use of the PyTorch or TensorFlow stack, as the role requires code that is readable and deployable.
Deepen knowledge of market microstructure
Research the basics of crypto derivatives and spot market microstructure. Familiarize yourself with order books, liquidity, and latency to better understand how your models might interact with real-world trading environments.
Formulate technical questions for the team
Prepare thoughtful questions regarding their backtesting infrastructure, how they handle model decay in live production, and the typical feedback loop between research and deployment to show your interest in the full lifecycle of trading models.
Refine communication of complex concepts
Practice explaining a complex ML model or statistical finding in simple terms. The role requires translating research insights for traders and engineers, so clarity and precision are essential.

Availability last checked: 19 Oct 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…