Nicolee D.

Key Skills

Software

No software listed

Top Subject Matter

Finance & Fintech

Logistics

IT

Top Data Types

Audio

Text

Video

Top Task Types

Transcription

Audio Recording

Prompt + Response Writing (SFT)

Data Collection

Computer Programming/Coding

RLHF

Freelancer Overview

My experience with AI training and data labeling comes from building my own AI systems from scratch. When I built Cura, my clinical intake platform for African healthcare settings, I had to make very deliberate decisions about how the model was being guided — structuring prompts, grounding outputs in curated clinical reference material, and deciding what the model should and should not be allowed to infer on its own. That process gave me a firsthand understanding of how the quality of input data and the structure of the retrieval layer directly shapes what a model produces, which is really what good data labeling practice is about at its core. On the learning side, my OCI Generative AI Professional certification gave me solid grounding in how models are trained, fine-tuned, and evaluated, including the role human feedback plays in shaping model behaviour over time. I have not held a formal data labeling role, but I have consistently been the person deciding what good output looks like, what counts as a failure, and how to structure information so the model produces something genuinely useful rather than just something that sounds right. That judgment — knowing the difference between a response that works and one that just looks like it does — is something I developed through building real products with real users, and I think it is the most honest version of AI training experience I can claim.

Education

Work History