Jean-francois Beaulnes

Key Skills

Software

Mindrift

Telus

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Top Task Types

Action Recognition

Audio Recording

Data Collection

Emotion Recognition

Fine Tuning

Freelancer Overview

My experience with AI training and data labeling spans more than three years while I have worked on Telus International and DataAnnotation and Outlier and Mindrift platforms. My work involved evaluating and rating AI-generated content in French (CA/FR) and English (CA) by assessing factual accuracy and prompt safety as well as fluency and localization. I have performed tasks that included prompt rewriting and response classification and fine-tuning to enhance LLM output quality. I designed a structured botanical database containing more than 40,000 plant entries for an educational website through my labeling work. The classification process involved sorting species according to taxonomic categories and ecological characteristics and regional distribution data through annotation and metadata tagging methods. My work experience includes domain quality evaluation and map-based Q&A review as well as "stump-the-model" biology tasks. My scientific background combined with bilingual skills and precise work methods enable me to produce top-notch annotations for both linguistic and science-based AI projects

Labeling Experience

Evaluating answers to prompts according to specific locale criteria.

TelusTextEvaluation Rating

For French(CA), French(FR) and English(CA)

For French(CA), French(FR) and English(CA)

2024

Writing prompts that will generate a human-like response to a voice query.

OtherAudioEmotion Recognition

Writing prompts according to criteria describing a human voice, notably, the pitch, the tone and the speed, the generate a human-like LLM vocal answer to a voice prompt.

Writing prompts according to criteria describing a human voice, notably, the pitch, the tone and the speed, the generate a human-like LLM vocal answer to a voice prompt.

2024

Image Data Extraction

OtherImageObject Detection

Extracting data, conceptual, and mathematical information from images.

Extracting data, conceptual, and mathematical information from images.

2024

Domain Quality Evaluation

TelusTextEvaluation Rating

Evaluating a wide range of websites according to well-defined criteria.

Evaluating a wide range of websites according to well-defined criteria.

2024

Assessing LLMs through biology problems

OtherTextQuestion Answering

Evaluating the chain of thought reasoning for a model. Then create a complex reasonning prompt that must stump the model in its chain of thought and final answer. You then have to rewrite the chain of thought.

Evaluating the chain of thought reasoning for a model. Then create a complex reasonning prompt that must stump the model in its chain of thought and final answer. You then have to rewrite the chain of thought.

2024

Show more

Education

Work History