Raghavender Reddy

- Migrated the legacy sustainability calculations engine to a vectorized dbt pipeline, reducing large-tenant processing time from 6 hours to 30 minutes. - Built real-time Change Data Capture (CDC) pipelines using PostgreSQL WAL, Kafka, and Snowflake, eliminating batch-sync latency.

- Architected the end-to-end data model and ETL integration for GRESB ESG reporting, driving 90% customer adoption. - Engineered an async FastAPI simulation microservice with Redis caching, enabling users to validate complex retrofit scenarios in seconds.

- Engineered a distributed inference pipeline using Celery and Redis to process, deduplicate, and analyze ~100 financial news articles per minute. - Established an end-to-end MLOps lifecycle using MLflow for experiment tracking, model versioning, and automated benchmarking.

- Developed a news recommender using FAISS and BERT, indexing 1M+ articles to deliver sub-2-second semantic search and market sentiment ranking.

- Scraped various websites and downloading the images and others metadata. - Trained convolutional neural network models, helps the users detect more than 300 dog and 50 cat breeds.

- Developed an algorithm that uses Kalman filters to detect the highway lane lines more accurately. - Trained convolutional neural network models to detect traffic signs, vehicles, and other moving objects. - Migrated monolithic application to microservice architecture.