At Burger Singh, I engineered an AP invoice data pipeline integrated with Business Central, combining dual-AI OCR, automated validation rules, and tiered approvals. The system achieved 93.1% model agreement.
I also built a reconciliation and payout automation system integrating data from four aggregator platforms, POS, and bank settlement feeds to validate deductions and automate downstream processing.
At 2Cents Capital, I built data architecture integrating more than five financial APIs with QuestDB and automated ingestion and validation pipelines. I also implemented portfolio risk analytics and evaluated quantitative strategies using walk-forward optimization and Monte Carlo validation.
In my projects, I analyzed New York State SPARCS data to compare machine-learning models for forecasting healthcare costs, and implemented a GLWB annuity pricing model combining GARCH and LSTM. I also developed a Python restaurant finder using a 2D Range Tree.
