Krishna Teja

Data Scientist @TechBlocks

Vancouver, BC, CA

Working with the Data Science team to build and scale forecasting and routing optimization models across North America- Contributing to the development of Hierarchical Bayesian models using PyMC to forecast weekly tank-level propane consumption- Designed and implemented MILP-based delivery optimization using Gurobi, accounting for delivery windows, driver capacity, and fuel constraints- Applied HDBSCAN clustering to spatially group customers for better route efficiency and training segmentation- Scaled forecasting pipelines to new regions (Atlantic, Mid-West.) using modular YAML-driven configs and automated validation- Engineered and monitored data workflows across Databricks (PySpark), Azure Synapse, and Blob Storage, enabling production-scale model training and inference- Automated downstream file delivery for BI teams using Azure APIs, and led efforts to decouple training datasets to improve pipeline resilience- Investigated kernel stalls in probabilistic models and collaborated on architectural improvements to ensure robustness and traceability.

Guru Nanak Institutions(GNI)

Bachelor of Technology - BTech