Rahul Joshi

- Developed tulanā's airline crew-scheduler from prototype to production: a stochastic optimizer in Python along with the data pipelines feeding it - Contributed to customer projects in inventory optimization and capacity planning: worked with customers to resolve data quality issues, built parts of the optimizer pipeline, and turned model output into KPI analyses and recommendations

- Built and presented end-to-end demos for blogs, conferences and talks, winning open-source users and technology partners for dltHub's open-core ELT library dlt - Set up internal pipelines, automations and reporting that scaled the company's CRM processes, replacing manual work across the go-to-market team - Worked extensively with the modern data ecosystem, ingesting from REST APIs and SQL databases into BigQuery, PostgreSQL, Snowflake, DuckDB, using dbt, Airflow, Metabase, Streamlit, and Google Cloud

- Delivered several data science projects for a variety of use-cases: anomaly detection, demand forecasting, NPV optimization, loan-default prediction, auto-cataloging - Extensively used Python and SQL to process, analyze, and model complex data from across different industry verticals (e-commerce, finance, manufactuing, oil & gas) - Effectively communicated technical results to both technical- and non-technical stakeholders

Focus: Machine Learning, Linear and Integer optimization (LP/MILP), Scientific Computing Thesis: on benchmarking solutions of large-scale Mixed-Interger programs (MILPs) of Berlin's district heating network