Mohamed Hassani

Key Skills

Software

Internal/Proprietary Tooling

Scale AI

Label Studio

Top Subject Matter

LLM, LLM Agents, Chatbots,

Data Manipulation

Object Detection

Top Data Types

Computer Code Programming

Document

Text

Top Task Types

Computer Programming Coding

Fine Tuning

Prompt Response Writing SFT

Question Answering

RLHF

Freelancer Overview

**AI Data & Quality Assurance Specialist** With a robust background in **AI data annotation, quality assurance, and model optimization**, I bring a unique blend of technical and analytical skills to enhance the performance of generative AI systems. At **Outlier**, I specialized in **Reinforcement Learning from Human Feedback (RLHF)** to improve large language models’ code generation and function-calling capabilities. My work included: - Designing and evaluating prompts to optimize model accuracy and functionality in code generation. - Providing detailed feedback and ranking AI-generated code (primarily in Python) to refine model outputs and practical applications. - Developing realistic programming simulations—such as bug fixes, code explanations, and modifications—to train models in responding to user requests effectively. - Implementing accurate code solutions and generating diverse code snippets to simulate real-world user interactions. - **Tech Stack:** Python, Pandas, PyQt6, FastAPI, OpenCV, Pygame. As a **QA Analyst at AdVon Commerce**, I focused on **data annotation and quality assurance for AI language models**, ensuring high-quality training datasets for LLM Agents. My responsibilities included: - Creating precise, structured training datasets while adhering to strict confidentiality and annota

Labeling Experience

Contractor Software Engineer

Scale AIComputer Code ProgrammingText GenerationRLHF

Enhanced large language models' code generation and function-calling abilities through Reinforcement Learning from Human Feedback (RLHF). Developed and evaluated prompts to optimize model effectiveness in generating accurate and functional code. Offered feedback and ranked AI-generated code (primarily Python) to improve model accuracy and practical application. Developed and implemented realistic programming simulation projects involving Python bug fixes, code explanations, and modifications based on user requests. Implemented accurate code solutions and modifications for simulated user requests. Generated code snippets and projects based on AI-simulated requests. Contributed to the improvement of generative AI performance across flexible projects in a remote setting. Tech: Python, Pandas, PyQt6, FastAPI, OpenCV, Pygame, ...

Enhanced large language models' code generation and function-calling abilities through Reinforcement Learning from Human Feedback (RLHF). Developed and evaluated prompts to optimize model effectiveness in generating accurate and functional code. Offered feedback and ranked AI-generated code (primarily Python) to improve model accuracy and practical application. Developed and implemented realistic programming simulation projects involving Python bug fixes, code explanations, and modifications based on user requests. Implemented accurate code solutions and modifications for simulated user requests. Generated code snippets and projects based on AI-simulated requests. Contributed to the improvement of generative AI performance across flexible projects in a remote setting. Tech: Python, Pandas, PyQt6, FastAPI, OpenCV, Pygame, ...

2024

QA Analyst

Internal Proprietary ToolingDocumentQuestion AnsweringText Generation

, I specialized in data annotation and quality assurance for AI language models. My experience aligns closely with the responsibilities of an Advanced AI Data Trainer, where I: • Created High-Quality Training Data: Generated precise and comprehensive training datasets to enhance the performance of LLM Agents models. • Reviewed Agents Outputs: Conducted thorough reviews of generated outputs to identify inaccuracies, improve error rates, and ensure the relevance of information. • Ensured Adherence to Annotation Guidelines: Strictly followed detailed annotation protocols to maintain consistency and accuracy in training data. • Maintained Confidentiality: Handled proprietary data and models with the utmost confidentiality to protect sensitive information. • Maximized Usefulness of Contributions: Deeply understood project goals to tailor my contributions effectively, enhancing the overall quality and usability of AI models.

, I specialized in data annotation and quality assurance for AI language models. My experience aligns closely with the responsibilities of an Advanced AI Data Trainer, where I: • Created High-Quality Training Data: Generated precise and comprehensive training datasets to enhance the performance of LLM Agents models. • Reviewed Agents Outputs: Conducted thorough reviews of generated outputs to identify inaccuracies, improve error rates, and ensure the relevance of information. • Ensured Adherence to Annotation Guidelines: Strictly followed detailed annotation protocols to maintain consistency and accuracy in training data. • Maintained Confidentiality: Handled proprietary data and models with the utmost confidentiality to protect sensitive information. • Maximized Usefulness of Contributions: Deeply understood project goals to tailor my contributions effectively, enhancing the overall quality and usability of AI models.

2023 - 2023

ML Engineer

Label StudioImageBounding BoxPolygon

Object detection models using open sources libraries like YOLO, DETR, MMdetection,..., from data preparation to model deployment.

Object detection models using open sources libraries like YOLO, DETR, MMdetection,..., from data preparation to model deployment.

2022 - 2023

Education

Work History