At Outlier AI, I design evaluation prompts for LLM alignment and recall testing, assess outputs for factual accuracy and reasoning, and rank model variants to contribute to fine-tuning improvements.
At Imperia Fincon Solutions, I built PySpark workflows for data cleansing and transformation and designed Power BI dashboards for KPI reporting. My projects include a data quality tool using the Anthropic API and fraud detection pipelines and dashboards.

