Open AI API & ReactJS April 1, 2025

ChatAI

This project is a custom AI driven assistant designed to manage real-time communication when you step away from your desk. The system acts as an autonomous responder within Slack, generating accurate and personalised replies by drawing on live context from your working day. It is built in ReactJS and powered by the OpenAI API, with optional support for a local LLM server. It connects directly to your workflow to understand what you are working on, who you have spoken to, and what your schedule looks like.

JavaScript Idle

React Idle

OpenAI Idle

Goal

The goal was to create an AI tool that does not guess but knows. Instead of generic automated replies, the aim was to build an assistant that responds with precision based on actual data such as your calendar, your tasks, active projects, time entries, emails, and previous conversations. Another important aim was to allow hands-off operation. You enter AFK mode and the system quietly takes over communication, keeping colleagues informed without interrupting your work.

Completion

The finished system integrates multiple data sources, processes them in real time, and produces natural and context aware responses that feel intentional rather than automated. It tracks working hours, reads your current workload, understands deadlines, and references past interactions without exposing sensitive information. The result is a reliable AFK communication layer that provides uninterrupted coverage while maintaining accuracy and professionalism. It is fast and efficient, thanks to ReactJS, careful API handling, and compatibility with cloud based and local language models.

ChatAI Video Example

AI Building

Constructing the AI layer required more than simple prompt engineering. The model needed a consistent picture of your day, your responsibilities, and your communication habits.

I built a structured context pipeline that pre-processes all relevant data before each message is generated. When Slack messages arrive, the system compiles a fresh context snapshot that includes your activities, availability, and ongoing work. This gives the model a clear grounding, reduces hallucination risk, and improves reliability. The AI then produces a response tailored to the person messaging you and the situation you are in.

Enable AI

Slack Clone

Integrating Calendar, Projects, Password App, Email, and Chat Data

A major part of the work involved safely connecting the app to key data sources. The system reads calendar entries to understand meetings, availability, and upcoming deadlines. It analyses project data to track what you are currently working on. It can securely reference password manager metadata to recognise accounts or services involved in conversations.

Email data helps it understand ongoing threads, priorities, and commitments. Existing chats, both active and historical, provide conversational context, tone, and continuity.