▸ OVERVIEW

ABOUT THE PROJECT

This AI chatbot is a general-purpose conversational assistant built from the ground up using Python and TensorFlow. Unlike simple rule-based bots, it uses machine learning models to understand natural language, infer intent, and generate contextually relevant responses.

The backend is powered by FastAPI, exposing a clean REST API that any frontend or integration can talk to. Conversations are tracked with a contextual memory system, meaning the bot remembers what was said earlier in the session and uses that context to give smarter, more coherent answers.

The goal is a fast, lightweight AI assistant that can be dropped into any project no third-party AI APIs required.

▸ INTERACTIVE DEMO

TRY THE CHATBOT

This is a live AI chatbot powered by the ChatGPT API. It has full contextual memory it remembers everything said earlier in your session. Ask it anything! *API is currently inactive*

8BP ASSISTANT

LIVE — CHATGPT API

🤖

Hey! I'm the 8BP Assistant, a general-purpose AI chatbot built by Riz. Ask me anything or try one of the suggestions below!

just now

▸ CURRENT STATUS

The core NLP pipeline, TensorFlow model architecture, and FastAPI backend are actively being developed. The GitHub repository and a hosted live demo will be published here once the project reaches a stable release. The interactive demo above shows the planned conversation style.

▸ FEATURES

WHAT IT DOES

🧠

NATURAL LANGUAGE PROCESSING

Understands free-form text input not just keywords. Tokenises, processes, and infers user intent using trained TensorFlow models.

💬

CONTEXTUAL MEMORY

Remembers prior messages within a session. Responses are informed by full conversation history, not just the latest message.

⚡

FAST REST API

Served via FastAPI async Python web framework with automatic OpenAPI docs, input validation, and sub-millisecond routing overhead.

🎯

INTENT CLASSIFICATION

Multi-class intent detection trained on a custom dataset. Classifies user queries into categories to route the correct response logic.

📦

GENERAL PURPOSE

Not locked to one domain. Designed to handle a wide range of everyday queries information, tasks, conversation, and more.

🔌

API-FIRST DESIGN

Clean JSON API means any frontend web, mobile, or desktop can integrate the chatbot with minimal effort. No vendor lock-in.

▸ TECHNOLOGY

TECH STACK

🐍

PYTHON

Primary language for all ML, NLP, and backend logic. Clean, readable, and the industry standard for AI/ML development.

CORE

🤖

TENSORFLOW

Powers the neural network model handles training, inference, tokenisation pipelines, and model serialisation for deployment.

ML ENGINE

⚡

FASTAPI

High-performance async REST API framework. Handles HTTP endpoints, request validation, automatic docs generation, and CORS.

API LAYER

▸ ARCHITECTURE

HOW IT'S BUILT

┌──────────────────────────────────────────────┐
│ CLIENT / FRONTEND │
│ Web App, Mobile, or API Call │
└──────────────────────┬───────────────────────┘
│ POST /chat {message, session_id}
┌──────────────────────┼───────────────────────┐
│ FASTAPI BACKEND │
│ ┌─────────────┐ ┌──────────┐ ┌─────────┐ │
│ │ ROUTER │ │ SESSION │ │ MEMORY │ │
│ │ /chat │ │ Manager │ │ Context │ │
│ │ /reset │ │ │ │ Store │ │
│ └──────┬──────┘ └──────────┘ └─────────┘ │
└─────────┼────────────────────────────────────┘
│ preprocess → tokenise → infer
┌─────────┼────────────────────────────────────┐
│ NLP PIPELINE (Python) │
│ ┌──────────────┐ ┌────────────────────┐ │
│ │ TOKENISER │ │ INTENT CLASSIFIER │ │
│ │ Text → Tokens│ │ TensorFlow Model │ │
│ └──────────────┘ └────────────────────┘ │
│ ┌──────────────────────────────────────┐ │
│ │ RESPONSE GENERATOR │ │
│ │ Intent + Context → Final reply │ │
│ └──────────────────────────────────────┘ │
└──────────────────────────────────────────────┘

▸ LINKS

EXPLORE THE PROJECT

GitHub and live demo will be linked here on first stable release.

⌀ GITHUB — COMING SOON ⌀ LIVE DEMO — COMING SOON

BACK TO

ALL PROJECTS

▸

AICHATBOT