Safaricom PLC
|Senior AI Training & Python Data Engineer
Summary
Led the architecture and implementation of programmatic data annotation workflows and rigorous evaluation frameworks for generative AI training pipelines at Safaricom PLC.
Highlights
Architected programmatic Python-based data annotation workflows, leveraging Pandas and NumPy to clean, validate, and structure massive datasets for generative AI training pipelines.
Designed and implemented complex Python scripts to rigorously evaluate LLM-generated code and text outputs against strict ground-truth benchmarks, successfully reducing error rates by 25%.
Authored comprehensive master annotation guidelines and gold-standard problem sets, standardizing edge-case classification across distributed AI training teams.
Developed automated Python reporting pipelines that aggregated data-labeling quality metrics from SQL databases, providing real-time analytics dashboards for performance monitoring.