Muhammad Y.

Key Skills

Software

CVAT

Snorkel AI

Other

Top Subject Matter

AI code review and LLM evaluation

LLM training/evaluation with SFT and RLHF

LLM orchestration

Top Data Types

Text

Computer Code Programming

Document

Image

Top Task Types

Prompt + Response Writing (SFT)

Data Collection

Function Calling

Bounding Box

Text Generation

RLHF

Freelancer Overview

AI Code Reviewer — Jupiter AI Labs. Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other, Internal, and Proprietary Tooling. Education includes Bachelor of Engineering, Federal University of Technology, Minna (2024). AI-training focus includes data types such as Computer Code, Programming, and Text and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

AI Code Reviewer — Jupiter AI Labs

Other

Performed AI code review and evaluation for correctness, logic, efficiency, readability, and adherence to programming best practices. Identified bugs, inconsistencies, and potential optimization opportunities, providing detailed feedback to improve both code quality and model performance. Conducted validation of AI-generated code outputs from large language model systems used in software development workflows. • Reviewed AI-generated code produced by models such as Anthropic’s Claude. • Evaluated outputs against engineering quality and robustness criteria. • Documented issues and recommended improvements to guide model refinement. • Focused on feedback loops to enhance downstream model behavior.

2025 - Present

AI Data Engineer — Technyder

OtherTextData Collection

Engineered LLM orchestration and retrieval-augmented generation pipelines to support agentic workflows in production settings. Delivered human-in-the-loop document processing and conversational analytics solutions intended to improve automated understanding of inputs. Built systems that included deep learning-based image analysis components for defect detection in marble slab production. • Configured RAG and agentic workflow components using LangChain. • Implemented human-in-the-loop processes for automated document handling. • Supported conversational analytics workflows for structured outputs. • Contributed to defect detection using deep learning image analysis pipelines.

2026 - 2026

AI Trainer — Turing

TextRLHF

Trained and evaluated large language models using supervised fine-tuning and reinforcement learning from human feedback approaches. Built evaluation tasks including function calling assessment, contextual reasoning checks, and prompt injection testing to measure instruction following and robustness. Developed prompts, multi-turn conversations, and edge-case scenarios to assess safety alignment, tool-use accuracy, and adversarial resilience. • Designed SFT and RLHF style training/evaluation scenarios. • Created multi-turn prompt sets and edge cases for model behavior testing. • Evaluated function calling correctness and context retention. • Performed red-team style prompt injection and manipulative input testing.

2025 - 2026

AI Data Engineer — AFEX

OtherDocumentData Collection

Designed, deployed, and maintained data infrastructure and business intelligence systems to enable reliable datasets for analytics and AI use cases. Supported orchestration and data integration workflows to keep staging environments consistent for downstream model development. Ensured datasets remained high quality and accessible for downstream AI and decision-making pipelines. • Built and maintained data pipelines on Ubuntu-based infrastructure. • Orchestrated and integrated data using Airflow and Airbyte. • Managed metadata and governance via OpenMetadata. • Enabled staging environments used by analytics and AI workflows.

2024 - 2026

Data Scientist (Freelance) — AI/ML Projects

OtherTextText Generation

Developed and optimized AI systems including generative RAG-based solutions for automated business proposal generation from RFP documents. Built and maintained retrieval-augmented workflows that used unstructured document inputs to produce structured generated outputs. Contributed to applied AI development tasks spanning computer vision, forecasting, and vulnerability-related log analysis. • Implemented RAG system for automated proposal generation. • Supported generative document workflows driven by RFP inputs. • Developed forecasting and measurement prediction models. • Built log analysis workflows for vulnerability detection.

2021 - 2024

Show more

Education

Work History