Rubeena Riyas

About

ML/data engineer who owned an end-to-end production sales analytics pipeline at 15,000+ user scale, delivering ~50% compute reduction, ~80% faster reporting, and ~$1.2M impact. Also shipped a production RAG-based AI assistant over internal BigQuery/docs with evaluation metrics and safety guardrails, and built shared Python libraries to standardize reliability and accelerate engineering teams.

Key Strengths

• Owned end-to-end production ML/data pipeline for sales analytics platform used by 15,000+ users
• Delivered major performance gains: ~50% compute time reduction and ~80% report speed improvement
• Drove ~$1.2M business impact through pipeline optimization and reliability improvements
• Deep expertise in data pipeline optimization (incremental processing, partitioning, predicate pushdown, Z-ordering, Spark tuning)
• Shipped production RAG assistant over internal sources (BigQuery + documentation) with evaluation and safety guardrails
• Balanced latency, cost, quality, and safety in production LLM systems (caching, smaller models, constrained queries, reduced tool calls)
• Built shared Python libraries with strong testing, versioning, and backward compatibility to improve team velocity and reliability

Reference Highlights

• Very hard worker
• Takes ownership and executes with limited requirements
• Strong at researching workflows and translating them into tools
• Delivers working toolsets that succeed with non-technical users
• Effective at diagnosing issues and shipping fixes quickly
• Able to distinguish code issues vs prompt/instruction issues
• Iterative improvement mindset (improved dozens of toolsets)
• Improved speed and productivity over time
• Deep education in ML/AI
• Maintained and improved critical toolsets over months
• Work had lasting customer impact (skills still working well today)
• Helped improve code review quality over time
• Good focus on reproducibility via tests and avoiding regressions

Experience

Education

Skills

Languages