AI / ML Engineer · Data Scientist · LLM Evaluation & Safety Research
Building data-driven systems, evaluating large language models, and designing robust ML pipelines with a focus on ethics, safety, and real-world impact.
- 🎯 Aspiring AI / Machine Learning Engineer with strong foundations in data analysis and systems thinking
- 🧠 Actively working on LLM evaluation, safety, bias, and benchmarking frameworks
- 📊 Experienced with Python, Pandas, NumPy, and structured datasets (JSONL, CSV)
- 🔍 Interested in model behavior analysis, robustness, and ethical AI
- 🚀 Long-term goal: contribute to reliable and transparent AI systems
- Python, SQL, Java (academic & projects)
- Pandas, NumPy, Matplotlib
- Scikit-Learn
- NLP preprocessing & data cleaning
- Model evaluation & benchmarking pipelines
- Safety / Ethics / Bias datasets (JSONL)
- Rubric-based scoring systems
- Model comparison & regression tracking (conceptual + implementation)
- Git & GitHub
- Jupyter Notebook
- Linux
- MySQL (integration with Python & PHP)
- 🧪 LLM benchmarking frameworks (safety, ethics, bias)
- 📐 Dataset schema design for evaluation tasks
- 🧠 Understanding reasoning, adversarial prompting, and failure modes
- 📊 Visualization & comparative analysis of model outputs
| Area | Description |
|---|---|
| LLM Evaluation Platform | Framework to benchmark LLMs on safety, ethics, and bias using structured JSONL datasets |
| Data Analysis Projects | Exploratory analysis, pivot tables, and visualizations using Pandas |
| NLP Data Cleaning | Annotation, preprocessing, and normalization of text data |
| Academic Projects | Java, assembly language fundamentals, and systems-level understanding |
📍 Only repositories where I am the original author are considered primary work.
- ✅ Prefer clean schemas and extensible designs
- ✅ Focus on evaluation, metrics, and behavior, not just accuracy
- ✅ Avoid unnecessary complexity (e.g., databases unless required)
- ✅ Build with future expansion in mind (reasoning, adversarial tests, compliance)
- Advanced NLP & LLM internals
- Reasoning and chain-of-thought evaluation
- Robust ML system design
- Research-level benchmarking methodologies
- GitHub: @MAqeel151214
- Open to collaboration on AI, ML, NLP, and evaluation research
"Build systems that can be trusted — not just systems that work."