Ngene Faith

Analytical Senior Data Annotator and Trainer with over 5 years of experience in enhancing large language models and computer vision systems. Expert in reinforcement learning from human feedback, multi-modal evaluation, and technical fact-checking. Proven ability to bridge the gap between complex human logic and machine learning outputs, with specialized expertise in advanced mathematics and linguistic analysis.

Lead AI Content specialist and Annotator - Outlier - Remote

(2021-01)

• Mode Optimization: Spearheaded high-priority RLHF initiatives, consistently ranking in the top tier of contributors for accuracy in complex reasoning tasks.
• Technical SME: Served as a subject matter Expert for technical domains including Electrical Engineering (project webster) and advanced Mathematical, verifying the integrity of AI-generated proofs and theorems
• Quality Control: Conducted rigorous Side-by-Side (SxS) evaluations for text and image models, identifying subtle hallucinations and ensuring adherence to safety guidelines.
• Workflow Management: Utilized remote monitoring and productivity tools (Hubstaff, AnyDesk) to maintain high-output performance across multiple long-term project lifecycles (Aether, Adelaide, Riverbed).
• Linguistic Evaluation: Applied deep understanding of syntax and grammar to evaluate model performance in specialized dialects and structured logic.

Data Evaluation Specialist - Cobra Clips & Various Projects - Freelance

(2021-04 - 2023-12)

• Visual Data Curation: Managed large-scale video and image annotation for the Cobra Clips dataset, improving model accuracy in object detection and temporal sequence recognition.
• Database Integrity: Categorized and labeled thousands of data points with 98.5% accuracy, adhering to evolving project taxonomies and strict deadlines.
• Feedback Loops: Collaborated with developers to refine annotation guidelines based on edge cases identified during the labeling process.

Independent Data Contractor - Remote Platforms - Remote

(2020-02 - 2021-03)

• Text Classification: Performed sentiment analysis and entity recognition for early-stage NLP (Natural Language Processing) datasets.
• Search Relevance: Evaluated search engine results for accuracy and user intent, contributing to the refinement of ranking algorithms.

B.Agric - Agriculture - Michael okpara university of Agriculture (2019-11 - 2024-05)