AI Benchmark Engineer | Native Language Specialist - Portuguese (Brazil) - Remote

LILT (Production) · Remote · Brazil (Remote) ·
Fully remoteContractMiddle3–5 years
COMPENSATION
€•• – ••k
0 applicants so far

Listed on the employer's job board (Ashby)

  • Native Portuguese (Brazil)
  • Fully remote freelance
  • Python + Shell scripting
  • AI benchmark creation
  • Flexible schedule

Join LILT as an AI Benchmark Engineer to build rigorous evaluation suites for Terminal-Bench tasks. You will create high-quality assets in Portuguese to measure how well AI models handle multilingual environments, focusing on robustness against prompt language effects and locale-specific challenges.

This is a remote freelance opportunity suited for supplemental income, offering flexible scheduling without fixed hours or micromanagement. You will collaborate with a global community of experts to advance multilingual AI capabilities.

What you'll do

  • Design realistic task environments using datasets and files in Portuguese
  • Identify failure points where AI struggles with native language prompts
  • Write deterministic verifier scripts and reference implementations
  • Calibrate task difficulty by analyzing execution logs across model tiers
  • Participate in a four-layer human quality control and audit process

What you'll need

  • Native or near-native fluency in Portuguese with deep grammatical knowledge
  • High proficiency in English for communication and documentation
  • At least one year of industry experience in software or prompt engineering
  • Strong skills in Python, standard shell scripting, and data processing
  • Extensive experience with Terminal/CLI-based development workflows
  • Understanding of Unicode normalization and encoding/decoding robustness

Nice to have

  • Background from leading technology companies or top-tier engineering universities
  • Familiarity with coding agents and their operational limits
  • Knowledge of locale-dependent conventions like collation and casing

Stack

  • Python
  • Shell scripting
  • Terminal-Bench
  • Coding agents

How you'll work

  • Fully remote freelance position (1099 contractor)
  • Flexible hours with no fixed schedule or check-ins
  • Requires reliable availability for 15-20 hours per week
  • Not suitable as a primary income source or full-time job
  • Contractors responsible for their own tax obligations

What you get

  • Competitive rates with prompt payment processing
  • Freedom to choose work volume and timing
  • Access to diverse AI and language technology projects
  • Opportunity to expand professional portfolio and skills

Hiring process

  • Submit application with an updated CV in English
  • Complete a GenAI assessment to evaluate technical skills
  • Finalize onboarding and profile setup in the system