Corporate Houston X53 Data Scientist

We are looking for experienced Data Scientist to implement ML initiatives. The responsibilities of the position include:

  • Build scalable data collection, transformation, integration pipelines.
  • Collect, clean, preprocess, and conduct FE.
  • Conduct statistical analysis and EDA to validate and extract interpretable insights from complex data patterns.
  • Select, train, validate, and serve models for generalized and specialized business use cases.
  • Implement fine-tunning and transfer learning pipelines with pretrained transformers.
  • Design, develop, and optimize end-to-end ETL and ML pipelines for production use.
  • Work closely with cross-functional teams to implement data-driven solutions.
  • Master's or Ph.D. degree in Computer Science, Statistics, Mathematics, or a related field.
  • Proficiency in Python, with hands-on experience in PySpark, Pandas, and PyTorch.
  • Expertise in data preprocessing, transformation, and manipulation techniques
  • Strong understanding of algorithms, including GBT, MLP/RNN/Transformer architectures.
  • Experience implementing model interpretability techniques.
  • Familiarity with big data platforms and distributed computing.
  • Strong communication skills, with the ability to translate complex, technical insights into clear, actionable recommendations for business stakeholders.
Show fields from Show fields from Show fields from a related table
Report Name *
Description
Reports and Charts Panel
Each table has a panel listing its reports and charts, organized in groups.
Please wait while your new report is saved...
Field label
Column heading override
Justification
What does auto mean?
Fields in:

Fields to Extract:
Name for the new table:
Items in the new table are called:

When you bring additional fields into a conversion, Quickbase often finds inconsistencies. For example, say you're converting your Companies column into its own table. One company, Acme Corporation, has offices in New York, Dallas and Portland. So, when you add the City column to the conversion, Quickbase finds three different locations for Acme. A single value in the column you're converting can only match one value in any additional field. Quickbase needs you to clean up the extra cities before it can create your new table. To do so, you have one of two choices:

  • If you want to create three separate Acme records (Acme-New York, Acme-Dallas and Acme-Portland) click the Conform link at the top of the column.
  • If the dissimilar entries are mistakes (say Acme only has one office in New York and the other locations are data-entry errors) go back into your table and correct the inconsistencies—in this case, changing all locations to New York. Then try the conversion again.

Read more about converting a column into a table.