Data Scientist / Quant Analyst

We are looking for an early-career Data Scientist / Quant Analyst who can turn messy market and positioning data into analysis we can trust and challenge it when we should not.

This is a hands-on role for someone who enjoys taking an unclear question, finding the evidence, and turning it into a defensible answer, model or analytical tool.

You’ll work closely with the Lead Data Scientist and the wider Data, Research, Product and Engineering teams. You will execute well, ask good questions, challenge weak assumptions, and earn larger ownership as your judgement grows.

About Vanda

Leaders in Positioning, Flow, and Tactical Macro

Vanda is an independent data and research firm delivering high-frequency positioning data and tactical macro insights. We combine proprietary positioning and flow data with market expertise to help investment professionals understand market set-ups, crowding, risk and opportunity.

Our data platform provides detailed views of investor positioning and flows, alongside Vanda research and advisory. We work across asset classes and geographies, with teams in London, New York and Singapore.

The Role

We are looking for an early-career Data Scientist / Quant Analyst to work closely with the Lead Data Scientist and the wider Data, Research, Product and Engineering teams.

This is a hands-on role for someone who enjoys taking an unclear question, finding the evidence, and turning it into a defensible answer, model or analytical tool.

You will not define the whole Data Science function from day one. You will execute well, ask good questions, challenge weak assumptions, and earn larger ownership as your judgement grows.

Key Responsibilities

• Solve quantitative problems: Turn scoped research, product or data questions into clear analytical tasks, baselines, tests and conclusions.
• Build and extend indicators: Reproduce and extend indicators across positioning, flows, equities, futures, options and cross-asset data.
• Prototype in Python: Prototype quickly, then turn useful work into clean, reusable and testable code.
• Test whether the story is true: Back-test and validate signals using time-series methods, out-of-sample checks, regime analysis and sensitivity tests.
• Challenge weak results: Look for leakage, look-ahead bias, unstable assumptions, overfitting, bad joins, survivorship issues, missing-data effects and weak charts.
• Speak up when the evidence is weak: Say when a result is weak, a model is too complex, or the data does not support the conclusion, even when the conclusion came from someone senior.
• Work with real, imperfect data: Profile large financial and alternative datasets, including coverage, history, identifiers, revisions, outliers, gaps, lags and structural breaks.
• Evaluate new data sources: Determine what they add, what they do not add, and whether the economics justify the complexity.
• Build diagnostics and QA: Create checks and monitoring that make analytical failures visible before they become product or client problems.
• Make research usable: Work with Data and Engineering to translate validated analysis into inputs, schemas, tests and production-ready implementations.
• Document your work: Clearly document methodology, assumptions, limitations and expected behaviour so others can reproduce and challenge the analysis.
• Support Research and Product: Contribute to market analysis and bespoke quantitative questions where rigorous data work improves the answer.