Software Engineer (Backend & Data) | Building and Scaling Systems | Python, SQL, JavaScript | CS @ IIT Patna

github

medium

Nisarg Gharde  • Peerlist

•Architected a fault-tolerant event streaming pipeline capable of buffering high - throughput gaming telemetry with zero data loss by deploying a Dockerized Apache Kafka (KRaft) message broker. •Optimized continuous time-series data ingestion and retrieval for live match history by designing a masterless Apache Cassandra NoSQL database with customized partition and clustering keys. •Served live gameplay analytics (such as top card meta and player snap rates) with sub-second latency by developing an asynchronous REST API using FastAPI and Pandas for in-memory data aggregation.

Marvel Snap Real Time Telemetry Pipeline

Engineered a customer churn predictor, achieving over 80% test accuracy by preprocessing 7,032 data entries and implementing a Random Forest classifier.

Telecom Operator Customer Churn Predictor

Developed a Python syntax validation tool using the compile() function to automatically analyze scripts and provide clear, line-specific error messages, improving code quality assurance.

Python Syntax Checker Tool

Built an end-to-end ML pipeline on Google Cloud Platform to predict customer churn for a telecom provider losing ~26% of customers annually. The platform ingests customer events in real-time via Cloud Pub/Sub, engineers churn-predictive features in BigQuery using defensive ELT patterns, and serves predictions through Cloud Functions loaded with a trained XGBoost classifier (ROC-AUC: 0.84+). Key engineering decisions: - Pub/Sub fan-out architecture — single topic feeds BigQuery, Dataflow layer, and Cloud Functions independently - Free-tier inference pattern — Cloud Functions + joblib artifact replaces Vertex AI endpoints at $0 cost - Feature parity between BigQuery SQL views and Python training pipeline ensures consistent scoring

Smart Customer Churn Intelligence Platform

Automated data quality analysis by engineering a Python and Pandas command-line interface (CLI) to extract structured insights, eliminate duplicate records, and export sanitized datasets for downstream processing.

Lightweight Data Profiling and EDA CLI Tool

Developed a zero-dependency Python syntax linter utilizing the native compile() function to statically analyze scripts and generate line-specific error reports, streamlining code quality assurance.

Lightweight Python Syntax Linter

Engineered a fault-tolerant, multi-cloud microservice in Go to process multipart file uploads and manage connection pooling to a Supabase PostgreSQL database. Implemented a true RAG architecture using Google Gemini 2.5 Flash for multimodal JSON extraction and Pinecone vector search to flag duplicate financial fraud attempts in real-time with 99%+ accuracy.

NexaAudit

Automated data quality analysis by building a command-line tool in Python and Pandas to deliver structured insights, remove duplicates, and export the cleaned data to a .CSV files.