At Invisible Technologies, I evaluate and annotate AI outputs across reasoning, factuality, safety, instruction-following, and multilingual tasks. I use detailed rubrics to produce structured feedback that supports reliable training data.
I work across high-volume multimodal workflows involving text, audio, images, translations, documents, PDFs, screenshots, and UI-based content. I also perform English and French transcription and voice evaluation, reviewing fidelity, intent, pronunciation, clarity, and naturalness.
Previously at Micro1, I assessed AI responses for accuracy, logic, policy issues, hallucinations, and edge cases, then rewrote outputs to improve clarity and real-world usability. I’m comfortable working independently in fast-moving remote projects with strict guidelines.
My customer support and training experience also shaped my approach to clear communication, QA, onboarding, and process improvement. I bring multilingual fluency in English, French, Swahili, Kinyarwanda, and Lingala to evaluation work that requires context and consistency.
