Available Contract Opportunities
Explore active freelance workstreams across frontier model training, multimodal labeling, and safety evaluation. Review duties, qualifications, and terms before you commit.
AI Model Training
Flexible Micro-tasks & MilestonesEvaluate and improve reasoning chains, multi-turn dialogues, and logical consistency in large language models.
Core Responsibilities
- Review prompt completions and rank response quality across standard rubrics
- Perform fact-checking and verify source citations in model arguments
- Identify reasoning breakdowns, logical leaps, or factual hallucinations
Target Qualifications
- Strong analytical reading and written communication skills in English
- Attention to nuance, factual precision, and logical coherence
- Reliable internet connection and modern desktop/laptop browser
Multimodal Data Annotation
Per-Task / Batch-Based PayoutsCategorize, segment, and label high-volume datasets across imagery, audio waveforms, video sequences, and structured text.
Core Responsibilities
- Apply polygon and bounding box annotations on visual datasets
- Transcribe and segment audio clips with phonetic or orthographic accuracy
- Tag syntactic and semantic entities in domain-specific corpus data
Target Qualifications
- High visual acuity and meticulous attention to visual boundary guidelines
- Ability to follow detailed multi-page annotation style guides
- Prior experience with data labeling tools is welcome but not required
Model Safety & Evaluation
Milestone-Based Hourly EquivalentStress-test models against safety guidelines, adversarial jailbreaks, bias detection, and toxic prompt handling.
Core Responsibilities
- Craft creative edge-case prompts to test safety boundaries
- Grade model outputs against toxicity, harm, and privacy rubrics
- Document failure modes and provide constructive mitigation feedback
Target Qualifications
- High ethical judgment and familiarity with AI safety principles
- Ability to engage with complex, potentially sensitive test material objectively
- Critical thinking skills and methodical documentation habit
Domain SME Evaluation
Premium Specialist CompensationDomain experts in STEM, law, medicine, finance, and software engineering providing specialized training datasets.
Core Responsibilities
- Generate and verify complex mathematical proofs, code solutions, or legal syntheses
- Grade advanced graduate-level AI prompt responses for technical accuracy
- Create challenging evaluation benchmarks within your specific discipline
Target Qualifications
- Documented background, degree, or professional track record in relevant field
- Deep technical mastery in code (Python/TS), advanced math, or specialized science
- Ability to explain intricate concepts with rigorous step-by-step reasoning
Agency Transparency Standard
All opportunities are subject to country eligibility verification, client platform agreements, and written terms. Northbridge does not charge application or placement fees. Actual compensation models, weekly hours, and assignment volume are finalized within the formal written project brief before any commitment is required.
Don't See Your Exact Background?
We frequently launch ad-hoc evaluation sprints for multilingual speakers, specialized coders, and domain researchers. Submit a general specialist profile to be considered for upcoming rosters.
Submit General Talent Profile