Fatma M.

I help teams verify AI-generated outputs, check research data, and catch mistakes before they reach users. My focus is accuracy, consistency, and work you can actually rely on. I've completed 450+ hours on Upwork across three projects, working on AI quality assurance, research, data verification, and Python-based data collection. My background is in Electrical Engineering, which has helped me develop a structured, detail-oriented approach to technical work. Recent Work: ✅ AI Agent QA – Wrote and tested instruction and skill files for an AI agent generating structured market research reports. Identified factual errors, broken logic, and inconsistent outputs, then revised the instructions to address those issues. ✅ Python Data Collection – Built scraping pipelines, including collecting 8,000+ X/Twitter posts, with cleaning and deduplication. ✅ Financial & Macro Research – Built and verified financial, economic, and crypto/DeFi datasets, with figures traced back to their original sources. What I Can Help With: ✅ AI Output QA & LLM Evaluation – Fact-checking AI answers, finding inconsistencies, scoring outputs against clear criteria, and spotting gaps in the workflows that produce them. ✅ Research & Data Verification – Checking facts, figures, references, and sources. ✅ Excel & Python Data Work – Collection, cleaning, deduplication, and validation. ✅ Technical Research & Documentation – Turning complex technical information into clear, structured deliverables. Why Work With Me: ✅ I don’t just flag something as wrong. I explain what’s wrong, why it matters, and what needs to change. ✅ I establish clear expectations and acceptance criteria before starting. ✅ I raise problems early instead of guessing. ✅ My findings are documented so they’re easy to understand and act on. Have an AI output, dataset, or research project that needs checking? Send me the details and a sample. I’ll tell you exactly what I’d check and how long it would take.