Designed and implemented a micro-services architecture using FastAPI, supporting both event-driven and REST-based communication with JWT authentication for AI-driven integrations.
Developed an RAG-based chatbot with conversational awareness, enabling students to interact with 100,000+ video and PDF-based contents stored in MongoDB as a vector database, using a Pub/Sub pattern for efficiency.
Implemented a semantic caching layer to optimise chatbot performance and reduce LLM calls.
Built an AI-powered assignment evaluator for automated grading and feedback.
Developed an automated content pipeline to generate summaries and interactive quizzes, mapping questions to timestamps and page numbers for enhanced learning experiences.
Created a dashboard in Angular to monitor AI-related costs and metrics.
Developed a reusable Python micro-services utility package, used across multiple services. Open-source version: https://pypi.org/project/zoldics-service-utils/
