William Gonzalez.

• Executed the core of company’s Alteryx-to-GCP migration: moved production workflows onto Airflow, dbt, and BigQuery with all code version-controlled in GitHub, retiring the legacy Alteryx estate and cutting annual platform spend from roughly $1.5M to $300-400K.
• Migrated transformations for BOOM, one of the company’s flagship client-processing programs, enabling 100+ clients to run fully automatically in Airflow and BigQuery instead of through manual Alteryx runs.
• Cut end-to-end client-processing latency from 6 hours to 1 hour for the full client portfolio, compressing the daily production window from 12 hours to 5 hours including built-in data-review checkpoints, with typical individual processes dropping from 3-4 hours to under 10 minutes.
• Standardized metadata-driven architectures that give non-technical teams 0-to-1 self-service speed: business users configure, launch, and modify complex data workflows from internal UIs without any engineering intervention or ad-hoc code changes.

• Conceived, built the proof of concept for, and led the company’s transition from legacy Alteryx and MySQL workflows to a cloud-native data ecosystem using Python, BigQuery, Airflow, PySpark, Terraform, Dataproc, and GCP services.
• Developed large-scale ETL pipelines orchestrated through Airflow, Dataproc, and PySpark to process terabytes of data daily.
• Reduced ROI reporting from 30 minutes to 20 seconds per report and from 8 hours to 2 minutes for company-wide execution.

• Conducted quantitative research with Bloomberg, FactSet, SEC filings and alternative financial datasets to support investment analysis.
• Produced investment research reports from proprietary analytical models and statistical insights.
• Developed statistical and predictive financial models in Python to evaluate corporate performance across multiple sectors.

• Researched machine learning applications for quantitative finance and hedging strategies.
• Modeled Deep Hedging with recurrent neural networks to evaluate quantitative hedging strategies.
• Evaluated predictive modeling approaches across structured and unstructured financial datasets.

Financial risk systems and reporting automation.

• Automated Reuters data retrieval — Python + Jupyter + SQL Server with Slack alerts, replacing manual workflows.~€60K saved annually
• Risk calculation systems in Java, Excel VBA, Wolfram Mathematica, Python — robust, scalable, audited.Production risk engine
• Eliminated repetitive financial ops — faster reporting cycles, lower error rate, more analyst time on analysis.Hours of manual work / week recovered

Banking infrastructure and reporting operations.

• Automated reporting cycle with VBA tooling for a major banking client.3 weeks → 1 day
• Project manager on the daily GTR project — primary liaison between bank managers and the IT team, weekly reporting to Executive Board.Owned client comms for largest GTR account
• Coordinated 8 bank departments to standardize income data from 14 sources into a unified ETL format.14 source systems unified