We are sharing a specialised part-time consulting opportunity for experienced software engineers with strong expertise in Python, Java, Rust, C++, Go, TypeScript, algorithms, debugging, refactoring, and performance optimisation to contribute to an advanced AI training project involving Model Context Protocol (MCP) environments. Selected professionals will create reinforcement-learning environments that test an AI model's ability to solve complex software-engineering problems using MCP tools and real server interactions. The work combines practical software engineering, deterministic evaluation design, and the creation of high-quality reference solutions. No prior experience in AI is required.
Key Responsibilities
- MCP Environment Development
- Create reinforcement-learning environments based on realistic software-engineering tasks
- Design scenarios requiring agents to discover and reason over information from MCP servers
- Build tasks involving real tool interactions rather than isolated code-generation exercises
- Ensure environments accurately measure both MCP tool use and engineering capability
- Maintain reproducibility across evaluation runs
- Software Engineering Task Design
- Create challenging tasks involving bug fixing, feature implementation, refactoring, and performance optimisation
- Develop scenarios that require meaningful reasoning across existing codebases
- Design tasks that test algorithms, data structures, debugging, and architectural judgement
- Ensure problems reflect realistic engineering constraints and workflows
- Balance task complexity with clear, measurable success criteria
- Golden Solutions & Deterministic Verification
- Create high-quality golden reference solutions for evaluation tasks
- Develop deterministic verification logic that reliably distinguishes correct from incorrect implementations
- Define clear acceptance criteria for software behaviour and task completion
- Validate environments against edge cases and unintended solution paths
- Ensure evaluation logic remains stable and reproducible
- Code Quality & Performance Engineering
- Debug complex software issues across multiple programming languages
- Implement maintainable features in existing codebases
- Refactor code while preserving intended functionality
- Identify and resolve performance bottlenecks
- Apply scalability, maintainability, and software-quality best practices
- Technical Review & Collaboration
- Review task quality, code correctness, and evaluation robustness
- Communicate technical decisions and assumptions clearly
- Participate in collaborative review of software-engineering environments
- Contribute to code-review standards and engineering best practices
- Work effectively in remote and cross-functional technical teams
Ideal Profile
- Strong proficiency in one or more of C++, Python, Java, Go, TypeScript, or Rust
- Deep understanding of algorithms, data structures, and performance optimisation
- Demonstrated experience debugging complex software issues
- Strong background in feature development and codebase refactoring
- Proven ability to improve software performance and scalability
- Experience working with large or distributed codebases is highly valuable
- Familiarity with rigorous code-review practices and software-engineering standards
- Strong written and verbal communication skills
- High attention to technical detail and reproducibility
- Experience with modern AI or machine-learning systems is beneficial but not required
- Prior AI-training or model-evaluation experience is not required
Engagement Details
- Part-time independent contractor engagement
- Fully remote
- Compensation: $60–$120/hour
- Expected commitment: approximately 15 hours per week
- Schedule is flexible, including the option to work evenings or weekends
- Compensation is output-based, with payment made for tasks that meet project specifications
- Minimum weekly submission requirements apply
- Work will involve MCP-based reinforcement-learning environments, software-engineering task design, deterministic verification, and golden reference solutions
- The selection process may include screening questions, an approximately 30-minute AI interview, a technical assessment, and hiring-manager review
- Selected professionals should be prepared to begin their first tasks within approximately 24–48 hours of completing onboarding
- Roles are typically filled within approximately 48 hours
- Work must be completed without using confidential or proprietary information belonging to any employer, client, institution, or other third party
About the Platform
- This opportunity is available through 24-MAG LLC. We connect experienced professionals with remote consulting opportunities across technical, evaluation, and project-based workstreams.
- By submitting this application, you acknowledge that your information may be processed by 24-MAG LLC for recruitment and opportunity matching in accordance with our Privacy Policy: https://www.24-mag.com/privacy-policy