Project Description
I’m building an AI-driven personal-assistant chatbot that runs on a Node.js + Express backend with MongoDB for persistence and Groq’s API for the language layer. The goal is to give users a fast, intelligent helper that works seamlessly on both web and mobile screens.
Scope of work
• Architect or refine the RESTful endpoints that receive user questions, forward them to Groq, and stream answers back in real time.
• Create (or polish) a responsive web interface and a companion mobile experience so the conversation feels native on any device.
• Persist every message, plus metadata needed for analytics, inside MongoDB and expose a history endpoint that can be filtered by user and date.
• Build two key features into the assistant itself:
– Task management: users can add, list, edit, and mark tasks done via natural language.
– Booking system: users can check availability, pick a slot, and receive confirmation, with data stored server-side.
• Optimise for speed and scalability—caching frequent prompts, async handling of Groq calls, and connection pooling for MongoDB.
Acceptance criteria
1. Average response latency (user request → first token) under one second with <200 concurrent users in test.
2. Task and booking commands processed accurately in ≥95 % of test cases.
3. All chat sessions retrievable by user ID with full pagination.
4. Front-end passes Google Lighthouse performance score ≥90 on mobile and desktop.
Code must be clean, documented, and delivered through Git. Please include a short README describing setup and any environment variables needed to run the stack locally.