AI Development
Advanced
Always open

Build a Cross-Language Code Translator

This challenge focuses on building an advanced AI agent system for high-fidelity code translation (e.g., Python to Go, Java to Kotlin). The system will leverage Gemini 2.5 Pro's strong code generation capabilities, specifically utilizing its 'Deep Think' mode for profound semantic understanding, and DSPy for programmatic prompting and optimization of the translation pipeline. The process will involve a multi-stage, graph-based workflow: first, understanding the source code's intent and structure; second, generating semantically equivalent target code; and third, robustly validating the translation. Validation will incorporate MCP-enabled static analysis tools, unit test generation, and potentially automated code review, ensuring the translated code is not only functionally correct but also idiomatic to the target language.

Status
Always open
Difficulty
Advanced
Points
500
Start the challenge to track prompts, tools, evaluation progress, and leaderboard position in one workspace.
Challenge at a glance
Host and timing
Vera

AI Research & Mentorship

Starts Available now
Evergreen challenge
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

This challenge focuses on building an advanced AI agent system for high-fidelity code translation (e.g., Python to Go, Java to Kotlin). The system will leverage Gemini 2.5 Pro's strong code generation capabilities, specifically utilizing its 'Deep Think' mode for profound semantic understanding, and DSPy for programmatic prompting and optimization of the translation pipeline. The process will involve a multi-stage, graph-based workflow: first, understanding the source code's intent and structure; second, generating semantically equivalent target code; and third, robustly validating the translation. Validation will incorporate MCP-enabled static analysis tools, unit test generation, and potentially automated code review, ensuring the translated code is not only functionally correct but also idiomatic to the target language.

Datasets

Shared data for this challenge

Review public datasets and any private uploads tied to your build.

Loading datasets...
Learning goals

What you should walk away with

Master DSPy for building programmatically optimized LLM pipelines, focusing on modularity and iterative refinement for code translation tasks.

Implement advanced prompt engineering techniques for Gemini 2.5 Pro, specifically utilizing its 'Deep Think' mode for profound semantic understanding and accurate, idiomatic code generation.

Design graph-based agent workflows (e.g., using a conceptual DAG for steps like AST parsing, intermediate representation, target code generation, and verification) for handling complex code structures.

Integrate MCP-enabled static analysis tools (e.g., linters, type checkers, code formatters) to verify the correctness, style, and idiomatic nature of the translated code.

Develop a feedback loop mechanism using DSPy's optimization capabilities to automatically improve translation accuracy based on static analysis results and unit test failures.

Build a comprehensive test suite that generates and executes unit tests for both source and translated code to ensure functional equivalence.

Explore techniques for handling advanced edge cases in code translation, such as complex dependency management, framework-specific constructs, and performance optimization.

Your progress

Participation status

You haven't started this challenge yet

Timeline and host

Operating window

Key dates and the organization behind this challenge.

Start date
Available now
Run mode
Evergreen challenge
Explore

Find another challenge

Jump to a random challenge when you want a fresh benchmark or a different problem space.

Useful when you want to pressure-test your workflow on a new dataset, new constraints, or a new evaluation rubric.

Tool Space Recipe

Draft
Evaluation

Frequently Asked Questions about Build a Cross-Language Code Translator