Senior ML/CV Engineer

Softeq; Polska; On-site; PLN 180,000 - 260,000; Full time

Job summary

Softeq, established in 1997, invites you to join a team delivering end-to-end product development and R&D. The role focuses on moving predictive models from experiment to production, building feature pipelines, and enabling reliable inference across environments.

You will work with Python, PySpark, and Databricks, maintain model endpoints, monitor latency and cost, and own release discipline. This is a hands-on senior ML engineering position in a growing global team, with opportunities to

Qualifications

• Production-grade Python and PySpark; Spark SQL proficiency.
• Databricks development & orchestration, job clusters, policies.
• Unity Catalog: grants, lineage, environment separation.
• Working Delta Lake knowledge: time travel, versioning.
• MLflow: experiment tracking, model registry.
• Feature storage and reuse, with consistent computation between training and inference.
• Hands-on experience operating models in production: monitoring, drift, retraining, incidents.
• Time-series and forecasting methods.
• Git, CI/CD, containers, secure handling of secrets.

Responsibilities

• Moves predictive models from experiment into regular operation: delay and duration forecasting, congestion forecasting, cost estimation, anomaly detection.
• Builds and maintains feature pipelines over the curated data layer, with a documented catalog of model inputs and their lineage.
• Makes training sets reproducible through table versioning and access to historical data states.
• Runs batch and near-real-time inference as platform jobs wired into the existing dependencies and schedules.
• Publishes models as endpoints for the application to consume, and watches latency and cost per call.
• Owns release discipline: versioning, retraining, rollback, promotion of models across environments.
• Provides quality monitoring, drift detection on inputs and outputs, input validation and explainability support.

Skills

Python

PySpark

Spark SQL

Time-series forecasting

Model monitoring

CI/CD

Git

Tools

Databricks

Unity Catalog

Delta Lake

MLflow

Containers

Azure

Job description

Established in 1997, Softeq was built from the ground up to specialize in new product development and R&D, tackling the most difficult problems in the tech sphere. Now we've expanded to offer early-stage innovation and ideation plus digital transformation business consulting. Our superpower is to deliver all of this under one roof on a global scale. So let's get started and build a better future together!

As we're expecting to expand our team and launch new projects within the next 1–2 months, we're already accepting applications and starting the interview process for selected candidates. We'd love to hear from you - feel free to apply!

What the role does here

• Moves predictive models from experiment into regular operation: delay and duration forecasting, congestion forecasting, cost estimation, anomaly detection;
• Builds and maintains feature pipelines over the curated data layer, with a documented catalog of model inputs and their lineage;
• Makes training sets reproducible through table versioning and access to historical data states.
• Runs batch and near-real-time inference as platform jobs wired into the existing dependencies and schedules;
• Publishes models as endpoints for the application to consume, and watches latency and cost per call;
• Owns release discipline: versioning, retraining, rollback, promotion of models across environments;
• Provides quality monitoring, drift detection on inputs and outputs, input validation and explainability support.

Must have

• Production-grade Python and PySpark, and confident Spark SQL;
• Databricks at development and operations depth: jobs and orchestration with dependencies, job clusters and cluster policies, portable project bundles, source control integration;
• Unity Catalog: grants on tables and models, lineage, environment separation through catalogs;
• Working Delta Lake knowledge: table versions and time travel, change data feed, optimization, and how data layout affects feature read performance;
• MLflow: experiment tracking, model registry, version promotion, publishing models as endpoints;
• Feature storage and reuse, with consistent computation between training and inference;
• Hands-on experience operating models in production: monitoring, drift, retraining, incidents, rollback;
• Time-series and forecasting methods;
• Git, CI/CD, containers, secure handling of secrets.

Nice to have

• Built-in platform monitoring for data and model quality;
• Production ML on Azure and integration with cloud services;
• Understanding of the platform consumption model and cost tuning for recurring jobs;
• Geospatial features and route data;
• Environments where a model must be explainable to business owners and auditors;
• Transportation, logistics or supply chain.

Softeq communicates only from @softeq.com email addresses. We never request payments or fees for any reason during hiring — including trainings or courses to be completed, equipment, onboarding, or background checks — and we will not ask for banking information, cryptocurrency or gift cards. If you receive a message from any other domain or requesting payment, do not respond and report it to [e-mail]