Skip to content

Instantly share code, notes, and snippets.

@zz2115
Created May 1, 2025 03:26
Show Gist options
  • Save zz2115/2d79b0132e0e834fbcf697cd46dddd84 to your computer and use it in GitHub Desktop.
Save zz2115/2d79b0132e0e834fbcf697cd46dddd84 to your computer and use it in GitHub Desktop.
Actions of a Great Data Scientist

Actions of a Great Data Scientist

Instead of simply copying notebooks, a great data scientist would:

  • Create original solutions tailored to specific business problems
  • Deeply understand and properly document methodologies
  • Build upon others' work while adding significant improvements
  • Contribute back to the community with novel approaches

Rather than training models blindly, they would:

  • Thoroughly explore and understand the data before modeling
  • Perform comprehensive exploratory data analysis
  • Identify and address data quality issues proactively
  • Document data lineage and characteristics meticulously

Instead of optimizing solely for accuracy, they would:

  • Balance multiple metrics relevant to business objectives
  • Consider trade-offs between accuracy, interpretability, and efficiency
  • Optimize for business impact rather than technical perfection
  • Develop solutions that address stakeholder priorities

A great data scientist would validate assumptions by:

  • Testing hypotheses rigorously before proceeding
  • Implementing systematic cross-validation approaches
  • Conducting sensitivity analyses on key parameters
  • Challenging their own biases and preconceptions regularly

Rather than avoiding stakeholders, they would:

  • Actively engage with business partners throughout projects
  • Translate technical concepts for non-technical audiences
  • Seek diverse perspectives to better understand requirements
  • Build relationships across departments to enhance collaboration

Instead of remaining stuck with basic tools, they would:

  • Continuously learn and adopt new technologies when appropriate
  • Move from exploration to production-ready implementations
  • Develop end-to-end solutions that deliver tangible value
  • Create reproducible and maintainable workflows

Rather than just delivering charts, they would:

  • Provide actionable insights with clear recommendations
  • Develop models that drive measurable business outcomes
  • Implement solutions that automate decision processes
  • Quantify the impact of their work in business terms

Instead of ignoring industry shifts, they would:

  • Strategically incorporate emerging technologies like LLMs
  • Balance innovation with practical implementation
  • Stay current with advances while focusing on business needs
  • Leverage new tools to solve previously intractable problems
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment