Are you interested in building machine learning systems that directly influence how digital products make decisions?
About the product
Forma Pro works with data-intensive products where machine learning is used to improve ranking, attribution, forecasting, anomaly detection, matching and operational decision-making.
Our projects can include marketplace monetization, advertising and attribution systems, partner and seller scoring, forecasting, anomaly detection and predictive features for operational platforms.
We are interested in ML systems that produce measurable business outcomes and can be continuously evaluated in production.
Technologies that we use
- Python
- Pandas and Polars
- NumPy
- scikit-learn
- PyTorch and modern ML libraries
- SQL
- PostgreSQL
- Jupyter
- ML experiment and evaluation tooling
- AWS
- Docker
- LLM APIs where they are useful
The exact stack varies by project. Choosing an appropriate approach is part of the job.
About you
- You enjoy translating unclear business questions into measurable ML problems
- You care about whether a model improves the actual product, not only whether it performs well in a notebook
- You are comfortable exploring imperfect real-world datasets
- You understand that a simpler model with good data can outperform an unnecessarily sophisticated solution
- You can work independently and communicate your reasoning clearly to engineers and non-technical stakeholders
- You are curious, pragmatic and willing to challenge assumptions
Required skills and competencies
- Spoken English — upper-intermediate or above
- 4+ years of professional experience in data science, machine learning or a closely related field
- Strong Python and SQL skills
- Solid knowledge of statistics and machine learning fundamentals
- Experience with supervised and unsupervised learning
- Experience with feature engineering, model validation and experimentation
- Understanding of data leakage, bias, drift and model performance monitoring
- Experience taking models beyond experimentation into production
- Ability to work with large and imperfect business datasets
Would be a plus
- Experience with recommendation or ranking systems
- Experience with attribution models
- Experience with anomaly or fraud detection
- Experience with forecasting
- Experience with marketplace, advertising or e-commerce datasets
- Experience with experimentation and A/B testing
- Experience with LLM evaluation or hybrid ML/LLM systems
- Familiarity with MLOps practices and model monitoring
Responsibilities
- Translate business and product problems into ML hypotheses and measurable objectives
- Explore datasets and identify useful signals
- Design, train and evaluate machine learning models
- Build ranking, scoring, forecasting, matching and anomaly-detection systems
- Define appropriate offline and online evaluation metrics
- Work with data engineers to improve training and inference datasets
- Work with backend engineers to integrate models into production applications
- Monitor model behaviour and identify degradation or data drift
- Design experiments to determine whether ML features create measurable product value
- Communicate model behaviour, limitations and trade-offs clearly to the wider team
- Research new methods when they can materially improve the system
What we offer
- No micromanagement — we trust experienced people to organize their own work
- Freedom to choose models, tools and approaches based on the problem rather than an internal technology checklist
- Flexible working hours
- Fully remote work
- Access to real production problems and datasets rather than artificial ML exercises
- The ability to influence both product decisions and technical architecture
- Company support when you need it
- Direct communication with Forma Pro's CEO and CTO