Cognitiv is looking to revolutionize the advertising industry by building the features that power their deep learning models, aiming to enable their Science team to build better models and drive better performance on campaigns.
Requirements
- Python
- SQL (e.g., ClickHouse)
- Spark
- C
- Docker
- RedPanda
- Flink
Responsibilities
- Build low-latency pipelines
- Deliver user features to our Feature Store within seconds of an event occurring, ensuring models have the freshest data possible.
- Leverage LLMs for embeddings
- Generate user and contextual embeddings that power advanced model performance.
- Maintain and expand features
- Ensure existing features are reliable and healthy while designing new ones to improve outcomes.
- Collaborate with Science
Other
- Hybrid work model (3 days in-office, 2 days remote)
- Collaborative partner – You thrive in cross-functional work and enjoy building with data scientists and other engineers
- Feature-focused mindset – You’re curious about how features influence models and motivated to improve them
- Machine learning exposure – Experience with ML pipelines or feature engineering
- Spark at scale – Hands-on expertise running large-scale jobs
