Python Quant Engineer — Where Research Becomes Alpha | AI-Native Systematic Trading

Mondrian Alpha; New York (NY); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Flat structure

Small teams

Direct access to traders

Job summary

Mondrian Alpha is seeking an exceptional Python engineer to build the platform powering iterative research and live trading. You will ingest petabytes of market data, train and serve models, and push signals into production safely and fast.

You will design high-performance backtesting infra, write kernels in Rust and C++, and collaborate with researchers, traders, and quants to ship tools that speed up research. The environment is flat with small teams and fast iteration.

Qualifications

• Strong Python engineering background and track record of optimization.
• Experience with large-scale data, ML infrastructure or quantitative research platforms.
• Willingness to drop down to Rust or C++ when Python isn’t enough.
• Strong CS fundamentals and a degree in a quantitative field.
• Engineer mindset treating research as a systems problem.

Responsibilities

• Build high-performance research and backtesting infrastructure that runs thousands of simulations in parallel across GPU and CPU clusters.
• Design columnar, zero-copy data pipelines on Arrow, Polars and DuckDB that handle tick-level history without friction.
• Write performance-critical kernels in Rust and C++ with PyO3/nanobind, and push hot paths through Numba, JAX and torch.compile.
• Productionise ML and LLM-assisted research workflows, from feature stores and model registries to low-latency inference.
• Work directly with portfolio managers and quants, shipping tools that change how they research.

Skills

Python engineering

CS fundamentals

Large-scale data experience

Rust/C++ interoperability

Systems thinking

Education

Quantitative field degree

Tools

PyO3 / nanobind

NumPy / Polars / JAX

CUDA / Triton

Job description

We are partnering with one of the world’s most technically ambitious systematic trading firms. They are rebuilding their research and execution stack for the AI era, with a single environment where ideas go from notebook to live markets in hours instead of weeks. The old split between researchers and engineers is gone, and the engineer who can span both is now the scarcest person in the market.

Their engineers work where research iteration speed and production reliability meet. Every second of backtest time saved is a day of research gained, and every model is measured in live PnL.

They’re now seeking an exceptional Python engineer to build the platform that powers that loop: ingesting petabytes of market data, training and serving models, and putting signals into production safely and fast.

The successful engineer will:

• Build high-performance research and backtesting infrastructure that runs thousands of simulations in parallel across GPU and CPU clusters
• Design columnar, zero-copy data pipelines on Arrow, Polars and DuckDB that handle tick-level history without friction
• Write performance-critical kernels in Rust and C++ with PyO3/nanobind, and push hot paths through Numba, JAX and torch.compile
• Productionise ML and LLM-assisted research workflows, from feature stores and model registries to low-latency inference
• Work directly with portfolio managers and quants, shipping tools that change how they research

The toolkit

• Modern Python: type-safe, async, free-threaded concurrency, and uv/ruff-grade engineering hygiene
• Vectorised and accelerated compute: NumPy, Polars, JAX, CUDA, Triton
• Profiling and observability: py-spy, Scalene, eBPF, OpenTelemetry
• A habit of benchmarking rather than assuming

Ideal background

• Strong Python engineering, with a record of making slow things fast
• Experience with large-scale data, ML infrastructure or quantitative research platforms
• Comfort dropping down to Rust or C++ when Python isn’t enough
• Strong CS fundamentals and a degree in a quantitative field
• The mindset of an engineer who treats research as a systems problem

The environment

Flat structure, small teams, no committees. Engineers sit alongside researchers and traders, and what you build is used by the people who trade on it the same week.

If you believe the next edge comes from better infrastructure for thinking, this is the place you’ve been building toward.