The company is hiring senior ML engineers to form the delivery core of its FinTech ML function.
Track A is Model Lifecycle. You run agentic model iteration covering experiments, evaluation, retraining and promotion. You also run agentic monitoring and observability for drift, stability and performance, built on the feature store. You design the agent loop and graph and keep it reliable and within cost.
Track B is Model Validation. You run agentic validation and backtesting. You build agent loops that replicate and stress-test models and produce evidence packs for independent reviewers. You also run agentic monitoring and observability for validation work.
The company is building its own fintech platform from scratch as a large international initiative. Engineers will shape the architecture, build core fintech components and deliver solutions meant to run at the scale of the wider group.
Responsibilities
Build and ship credit-scoring and merchant-underwriting models from start to finish, covering feature engineering, training, validation, deployment and monitoring after launch.
Create point-in-time-correct features from the shared feature store for BNPL and MCA decisions. Work with Data Engineering on source contracts and data freshness targets.
Design and run the agent loop and graph, including context, tools, evals and guardrails, so agentic delivery stays reliable, observable and affordable.
Deploy and operate models using MLOps tooling for versioning, model CI/CD, and champion/challenger or shadow rollouts.
Watch production models for drift, instability and falling performance. Retrain, recalibrate or roll back when thresholds are breached.
Build models that are explainable and defensible to regulators from the start. Document features, decisions and validation evidence for CCD2 creditworthiness checks and audits.
Validation track: build and run the validation and backtesting harness within the MLOps pipeline, and produce audit-ready evidence packs for independent review.
Lifecycle track: build the loop for experiments, evaluation, retraining and promotion on the model registry and serving layer.
You will work with the feature store, model registry and serving, agent orchestration graphs, eval harnesses, monitoring and tracing tools, the validation and backtesting harness in the MLOps pipeline, and audit-ready evidence outputs.
Requirements
You enjoy building ML systems that hold up in production and under audit.
You build hands-on and use AI tools every day.
You have applied ML experience at scale across the full lifecycle: feature engineering, training, validation, deployment and monitoring.
You know model validation methods well, including backtesting, population stability index (PSI), discrimination measures (Gini/KS) and calibration.
You have experience with MLOps and production monitoring, including model versioning, CI/CD, champion/challenger and shadow deployment, drift detection and automated alerts.
You can build point-in-time-correct features and consume feature stores properly.
You understand explainability methods such as SHAP and LIME and can write documentation that supports regulatory defensibility.
You write strong Python and SQL, and use Spark/Databricks for large-scale training and serving.
You have advanced experience with agentic tooling and with loop and graph engineering, including context and tool design, guardrails, and evals with auditable outputs. You also know LLMs well.
You work well with others, follow shared patterns and raise the standard through influence.
You speak English fluently (B2+).
Bonus: hands-on model validation or independent-review experience.
Bonus: CCD2 and consumer-credit rules for creditworthiness assessment.
Bonus: merchant cash advance (MCA) underwriting and merchant-risk data.
Bonus: fraud-detection modelling, such as bust-out and first-party fraud.
Bonus: IFRS 9 provisioning or expected-credit-loss modelling.
Bonus: real-time, low-latency model serving.
Conditions
Hybrid work, 1 day a week in the Warsaw or Kraków office.
B2B contract.
Benefits
A new build where the stakes are high.
Influence over how the company approaches AI-native ML.
A modern stack and a lean, AI-native team. You have room to test new technologies, including AI tooling, and move them into production.
A fast-moving, collaborative setting where autonomy and accountability go together.
The company is growing a lean, effective engineering team in Poland with AI-native engineering at its core. It is looking for engineers comfortable with agentic work, from experienced builders of complex systems and tooling to newer engineers who have worked with AI-driven development from the start of their careers.