About the Role:
At C4Scale, we are building AI-powered financial intelligence systems that bring together company fundamentals, market data, and information from documents and external sources.
We are looking for a hands-on engineer who has built and evaluated predictive models for stocks, investment research, or closely related financial applications. You should understand both the modelling challenges and the engineering needed to turn research into a reliable product.
You will own the work from data preparation and hypothesis development through model evaluation, deployment, and monitoring. This includes improving existing document extraction pipelines and exposing model outputs through APIs.
This role suits someone who is deeply curious about financial markets, takes initiative, and can independently turn an ambiguous question into a measurable engineering outcome.
What You Will Do:
Build financial prediction models: Develop models for stock return probabilities, downside risk, and related financial outcomes using company fundamentals, market behaviour, and external information.Develop meaningful features: Translate financial statements, business characteristics, sector context, and unstructured information into features with a clear economic rationale.Evaluate models rigorously: Build reproducible backtests and walk-forward evaluations. Address look-ahead bias, survivorship bias, overlapping prediction windows, and changing market conditions.Measure practical value: Evaluate probability calibration, prediction coverage, false positives, drawdowns, and performance after transaction costs. Test risk and exit policies separately from predictive accuracy.Improve extraction and data pipelines: Use LLMs and deterministic validation to extract structured, source-grounded facts from documents. Handle entity matching, missing data, duplicates, and data-quality failures.Ship production services: Build maintainable Python pipelines and APIs for model inference, supporting evidence, and confidence estimates.Own continuous improvement: Track experiments, data and model versions, inference costs, and production drift. Communicate findings, limitations, and next steps clearly.What You Will Need:
Demonstrated experience building and evaluating stock prediction, quantitative research, investment analytics, or closely related financial ML models. Be ready to explain your own contribution, evaluation design, and results on unseen data.Strong Python and SQL skills, with the ability to write modular, tested, production-ready code.Solid foundations in probability, statistics, supervised learning, time-series analysis, feature engineering, and model evaluation.Practical experience with pandas or Polars, NumPy, scikit-learn, and a gradient-boosting library such as LightGBM or XGBoost.Understanding of company fundamentals and financial statements, alongside returns, volatility, liquidity, corporate actions, and market benchmarks.Experience constructing datasets using only information available at the prediction time, including correct handling of publication dates and subsequently revised data.Hands-on experience with LLM APIs, structured extraction, schema validation, and evidence-based evaluation. Ability to measure extraction errors independently from downstream model performance.Working experience with PostgreSQL, FastAPI or an equivalent Python API framework, Git, Docker, and automated testing.High ownership: you identify missing information, investigate failures, propose sensible experiments, and carry work through to completion with limited day-to-day direction.Preferred Qualifications / Added Advantage:
Models deployed into production or evaluated prospectively through paper trading or a comparable live research process.Experience combining financial text and numerical data using NLP, embeddings, or transformer models.Familiarity with uncertainty estimation, model explainability, and deciding when a model should abstain.Experience with MLflow or equivalent experiment tracking, scheduled data workflows, and AWS deployment.Practical use of PyTorch where deep learning provides a demonstrated benefit.A relevant project, research publication, open-source contribution, or anonymised case study showing how you tested a financial hypothesis and what you learned when it failed.What We Offer:
Ownership of a challenging financial intelligence product, with direct access to the founder and technical leadership.The opportunity to shape modelling decisions and take systems from research into production.A macOS laptop.Competitive salary and benefits.A collaborative environment that values initiative, technical depth, and honest measurement.Continuous learning and professional development opportunities.