AI Sentiment Analyser
Real-time sentiment analysis of customer reviews using fine-tuned BERT with a FastAPI backend and React dashboard processing 100K+ reviews daily.
View project →A complete catalogue of systems, tools, and products I've built across machine learning, backend infrastructure, and data engineering.
Real-time sentiment analysis of customer reviews using fine-tuned BERT with a FastAPI backend and React dashboard processing 100K+ reviews daily.
View project →LSTM-based time-series forecasting model with Monte Carlo dropout confidence bands, deployed serverlessly on AWS Lambda for under 200ms predictions.
View project →End-to-end MLOps pipeline for churn prediction with automated retraining, data drift detection, and Grafana monitoring — built on MLflow, Prefect, and XGBoost.
View project →Retrieval-Augmented Generation system for private document Q&A using LangChain, FAISS vector search, and GPT-4 — supporting conversational follow-up questions.
View project →Scalable Kafka → Spark Streaming → ClickHouse pipeline ingesting 100K+ events/second, visualised in a self-service Apache Superset dashboard on Kubernetes.
View project →Python automation tool that analyses AWS and GCP billing APIs, identifies waste (idle instances, unused EBS, orphaned snapshots), and auto-generates Terraform remediation scripts.
View project →A backend service that hooks into GitHub Webhooks and runs GPT-4 to post line-by-line code review comments on every pull request, with smart deduplication and configurable severity rules.
View project →