Help shape the future of multilingual AI by designing and validating benchmark tasks that push large language models beyond English-centric workflows. In this role, you’ll engineer realistic coding challenges in Serbian, ensuring test assets remain in the native language to accurately assess a model’s ability to process non-English inputs without translation fallbacks.
Key Responsibilities
- Develop and refine coding tasks that expose weaknesses in AI handling of Serbian language inputs within terminal-based environments
- Create authentic data sets and file structures in Serbian to simulate real-world multilingual software workflows
- Design prompts that identify failure points in AI reasoning when processing native-language commands and text
- Build deterministic verification scripts and support reference implementations to validate model outputs
- Calibrate task difficulty across model tiers using standardized evaluation protocols
- Participate in a multi-stage quality assurance process combining human review, peer calibration, and automated checks
Required Expertise
- Five or more years of professional software engineering experience
- Demonstrated background at top-tier tech firms or institutions
- Native or near-native fluency in Serbian, including mastery of grammar, register, and linguistic nuance
- Strong written and spoken English proficiency
- Proficiency in Python, shell scripting, and data manipulation tools
- Extensive experience with command-line interfaces and coding agent workflows
- Deep knowledge of multilingual text processing challenges, including encoding, Unicode normalization, locale-specific formatting, and bidirectional text handling
- Familiarity with text I/O safety, toolchain interoperability, and rendering edge cases in non-English contexts
Work Environment
As an independent contractor, you set your own pace and availability. There are no fixed hours or mandatory check-ins—just meaningful work with full autonomy. You’ll contribute to cutting-edge AI evaluation frameworks while joining a global network of language and technical experts. Payments are competitive and processed promptly, with no administrative delays.
Commitment to Fairness
We uphold a hiring process grounded in inclusion, transparency, and equity. Our practices are designed to eliminate bias and ensure equal opportunity for all applicants, regardless of race, gender, religion, age, disability, national origin, or veteran status. This role supports a broader mission to advance human-machine communication across languages and cultures.
