AI & ML Skills

Skills for AI and ML workflows — prompt engineering, RAG building, model evaluation, fine-tuning, and AI safety checklists.

Iteratively refines prompts using chain-of-thought, few-shot examples, and role framing to maximize LLM output quality and consistency.

Scaffolds retrieval-augmented generation pipelines — chunking strategies, embedding selection, vector store configuration, and retrieval tuning.

Runs structured evaluation harnesses — accuracy, hallucination rate, latency benchmarks, and BLEU/ROUGE scoring across model versions.

Generates annotation schemas, labeling guidelines, and quality-control checklists for supervised ML datasets across text, image, and tabular data.

Advises on fine-tuning strategies — dataset size requirements, hyperparameter tuning, LoRA configuration, and overfitting prevention techniques.

Audits AI system outputs for bias, toxicity, and hallucination patterns — generates safety reports with remediation recommendations per use case.