Rust’s type system for data engineering: if it compiles, your data Is correct

🧱 Workshop overview

Every data engineer has shipped a pipeline where a column name was misspelled, a type was silently coerced, or a nullable field failed in production.

In Python and Spark, these are runtime surprises.
In Rust, they can be compile-time errors.

This hands-on workshop shows how Rust’s type system can encode data contracts - schemas, column types, valid states, and query structure - so the compiler catches bugs before your pipeline runs.

Participants will build a type-safe mini analytics pipeline from scratch using patterns from production systems like DataFusion and Polars. Each concept is introduced through short, incremental exercises-ending with a system where:

If it compiles, the data is structurally correct.

Workshop level: intermediate

🎯 Key learnings

By the end of this workshop, participants will be able to:

• Use the newtype pattern to prevent mixing domain types (e.g., CustomerId vs OrderId)
• Apply phantom types to enforce data lifecycle stages (raw → validated → transformed)
• Design trait-based transform pipelines where incompatible schemas fail at compile time
• Build typestate-driven query APIs that prevent invalid execution paths
• Understand how systems like DataFusion and Polars use these patterns internally

💡 What you’ll build

A working type-safe analytics pipeline, built incrementally using:

• Newtypes → Prevent invalid domain mixing
• Phantom types → Encode data lifecycle stages
• Trait-based transforms → Enforce schema compatibility
• Typestate + Builder patterns → Guarantee valid query construction

🚀 Takeaways

• Four production-grade Rust patterns:

  • Newtype
  • Phantom types
  • Trait-based transforms
  • Typestate

• A working pipeline you built yourself, demonstrating how to:

  • Encode correctness at compile time
  • Eliminate entire classes of runtime data bugs

👥 Target Audience

• Rust developers building data-intensive systems
• Data engineers exploring Rust
• Engineers frustrated with runtime schema/type errors in pipelines

👩 🏫 / 👨 🏫 About the instructors

Navdeep Agarwal & Mayur Jadhav are co-founders at OrcaSheets. They build squeaking fast local first analytics engine.

🎟️ How to Attend

This is a hybrid workshop — you can attend:

• In-person in Bengaluru, or
• Remotely (live online)

To participate, please purchase a workshop ticket for 17 April.

Note:

• Conference tickets do not include workshop access
• A valid workshop ticket is required for both in-person and remote participation
• Seats (both in-person and remote) are limited

📞 Contact

• Email: info@hasgeek.com
• Phone/WhatsApp: +91-7676332020
• Questions: https://hasgeek.com/rustbangalore/rusts-type-system-for-data-engineering/comments