Freelancer profile translated to English.

Description

I am David, an independent consultant in applied artificial intelligence, specializing in the improvement of language models (LLMs).

I work on RLHF, prompt engineering, data annotation, and quality assurance (QA) missions, with a focus on the reliability and performance of systems in real-world conditions.

My work involves analyzing and improving the quality of data and responses generated by models, by participating in dataset structuring, AI output evaluation, and prompt optimization.

I pay particular attention to the consistency of results, the detection of errors or biases, and a fine understanding of model behaviors.

I have a very rigorous approach to work, focused on the quality and concrete usefulness of the improvements made. The goal is to contribute to AI systems that are more reliable, more stable, and better aligned with real-world uses.

Languages

French
Native or bilingual
English
Fluent

Workplace preferences

Remote only

Primarily works remotely

Scale AI
AI / Data Annotation / Prompt Engineering
E-COMMERCE
January 2025 - March 2025 (2 months)
Toulouse, France
- Annotation and labeling of textual data for training conversational AI models.

- Evaluation of the relevance and safety of responses generated by AI assistants.

- Quality assurance (QA) of datasets before integration into training pipelines.

- Creation and cleaning of datasets for fine-tuning NLP models.

- Writing prompts for customer service and e-commerce assistance.

- Participation in simple RLHF tasks (ranking AI responses by quality, accuracy, and tone).

Skills required for this mission

Data Annotation · RLHF · Prompt Writing · AI QA · Dataset Cleaning · NLP
Applied Artificial Intelligence (LLM) Data Annotation Prompt engineering
Telus International AI
Junior AI Evaluator
HEALTH AND WELLNESS
April 2025 - June 2025 (2 months)
Toulouse, France
- Testing and rating of generative AI models used in non-sensitive medical contexts.

- Checking consistency, accuracy, and safety of automated responses.

- Comparison of AI outputs based on different criteria: precision, clarity, and neutrality.

- Creation of dialogue scenarios for conversational evaluation.

- Annotation of multilingual data (French/English).

- Collaboration with quality teams to improve evaluation guidelines.

Skills for this mission
AI Evaluation · Quality Assurance · Dialogue Testing · Data Labeling · Multilingual QA
Outlier AI
Prompt Engineering Assistant
EDUCATION AND E-LEARNING
September 2025 - December 2025 (3 months)
Toulouse, France
- Design of text and image prompts to test educational AI models.

- Creation of simulated conversations for training pedagogical AI agents.

- Verification of hallucinations and inconsistencies in generated responses.

- Structuring clean datasets for reading comprehension tasks.

- Optimization of prompts to improve response relevance.

- Participation in improvement cycles