Data Scientist - Feature Engineering

The Voleon Group•New York City, NY

21h•$150,000 - $190,000

About The Position

Voleon is a technology company that applies state-of-the-art AI and machine learning techniques to real-world problems in finance. For nearly two decades, we have led our industry and worked at the frontier of applying AI/ML to investment management. We have become a multibillion-dollar asset manager, and we have ambitious goals for the future. Your colleagues will include internationally recognized experts in artificial intelligence and machine learning research as well as highly experienced finance and technology professionals. In addition to our enriching and collegial working environment, we offer highly competitive compensation and benefits packages, technology talks by our experts, a beautiful modern office, daily catered lunches, and more. The Voleon Group is growing its Feature Engineering team — a small, high-impact group responsible for turning the world's messy, complex datasets into predictive signals that power our machine learning models. As a Data Scientist on this team, you'll dig into raw data from diverse domains and, in collaboration with Research, assist in design and implementation of features that capture what is actually happening in the world. This is data storytelling at its most consequential: every feature you build has a direct path to our investment process. We're looking for deeply curious people who get genuine satisfaction from wrestling with an unfamiliar dataset, understanding its structure and quirks, and emerging with something that encodes real information. You'll own the full arc from data sourcing and curation through feature construction, statistical validation, and integration into our production systems. We also expect you to actively experiment with AI-powered tools — LLM-based coding a ssistants, agentic workflows, and whatever comes next — to accelerate your day-to-day work and push the boundaries of what a small team can accomplish.

Requirements

2 years of applied in dustry experience (including internships) working end-to-end with complex datasets: curation, querying, aggregation, exploratory analysis, and visualization
Experience using statistical methods to analyze data, identify patterns, conduct root-cause analysis, and translate findings into actionable insights
Ability to frame and answer questions mathematically
Ability to infer useful forward-looking directions from the results of retrospective analysis
Fluency in managing, processing, and visualizing tabular data using SQL and Python (Pandas or Polars)
Basic software development skills and experience with bash, Linux/Unix, and git
Ability to refine ambiguous requests into well-scoped analyses and communicate results with clarity and precision
Bachelor's degree in a quantitative discipline (statistics, data science, computer science, economics, physics, or a related field)

Nice To Haves

Master's degree in a quantitative discipline
Prior industry experience or demonstrated interest in finance — academic projects, coursework in financial engineering, or industry internships
Familiarity with financial datasets such as Compustat, IBES, or similar vendor data
Experience developing in a production-facing environment with standard tooling (CI/CD, git, workflow orchestration)
Hands-on experience with AI coding assistants or LLM-based tools in a data science or engineering workflow
A track record of curiosity-driven exploration — side projects, Kaggle competitions, research papers, or anything that shows you can't leave an interesting dataset alone

Responsibilities

Explore, profile, and curate complex and often messy datasets from third-party vendors and internal sources, developing a deep understanding of what each dataset can and cannot tell us
Harness financial intuition, academic research, and statistical rigor to inform design and implementation of predictive features in collaborative setting
Validate features through a disciplined, test-driven framework — including cross-sectional analysis, stationarity testing, and point-in-time correctness — to ensure signals are real and not artifacts of data issues
Build and maintain data pipelines that bring features from prototype to production, with monitoring for data health and correctness along the way
Communicate your findings clearly — both the signal you've found and the story of how the data produces it — to researchers and leadership
Proactively investigate anomalies in data feeds and production behavior, performing root-cause analysis and surfacing issues to relevant stakeholders
Leverage AI tools to accelerate exploration, coding, and analysis — and share what you learn about effective workflows with the team