Role overview
We are seeking a highly skilled Data Scientist to design and develop pipelines and AI/ML models and workflows on our 50+ alternative data panels. The ideal candidate will have deep expertise in mathematical and statistical modeling, and will have built models with Python, SQL, and PySpark. This person will test new data assets for the firm, develop AI/LLM tools and agents, contribute to the firm’s analytics library, and use traditional machine learning and statistical methods to improve panel data. M Science expects its data scientists to implement production code, so the ideal candidate will have experience writing well tested, performant, object-oriented code.
What you'll work on
- Develop agentic workflows for automated insight retrieval, data analysis, data download, and data forecasting
- Contribute to the firm’s documented, unit-tested analytics library
- Process, cleanse, and verify the integrity of data used for analysis
- Create automated alerting and notification systems for deviations in data quality, validation failures, or unusual patterns
- Evaluate new datasets for the firm
- Design, develop, and optimize scalable and fault-tolerant data ingestion pipelines using Databricks, Airflow, Python, and Spark
- Build resilient data pipelines that handle vendor-related issues such as delayed deliveries, schema changes, incomplete records, and data corruption
What we're looking for
- Advanced Python for data processing, scripting, and automation
- Fluency in PySpark/Spark and distributed data processing; pandas, dask, daft, polars also a plus
- Excellent knowledge of multivariate statistical analysis, including but not limited to ordinary least squares, principal component analysis, factor analysis, LDA, and panel methods
- Excellent knowledge of other ML methods including additive modeling and ensemble modeling
- Some knowledge of LLMs and common LLM orchestration frameworks like LangChain and LangGraph
- Experience with named entity resolution methods a strong plus
- Familiarity with cloud data platforms (AWS) and cloud-based storage solutions
- Strong troubleshooting skills to diagnose and resolve performance bottlenecks in data pipelines
Tags & focus areas
Used for matching and alerts on DevFound Fulltime Data Science Ai