Live · RAG-powered · Logfare AI · Free for students

Train your own private AI
from PDFs & notes — chat like ChatGPT.

Upload a textbook, lecture notes, or paste any text. SmartTrain AI uses RAG (Retrieval-Augmented Generation) + Logfare AI to answer questions accurately from your data, and chat naturally for general questions. Built by Sarc Education Foundation to help students understand how AI really works.

Create your AI workspace Resume with Access ID See how it works
smarttrain.ai/chat/AX92KD
What is photosynthesis?
Photosynthesis is the process by which green plants use sunlight, water, and carbon dioxide to make glucose and release oxygen. It happens mainly in the chloroplasts of leaves.
Hi, how are you?
Hello! 👋 I'm doing great and ready to help. Ask me anything from your training data, or just chat!

Everything you need to build & chat with your AI

Lightweight. Modern. Mobile-first. Works on any shared hosting.

PDF + RAG indexing

Auto-extracts text, chunks it with overlap and builds a searchable vector index.

Logfare AI

Natural conversation, reasoning and answer composition powered by Logfare AI.

Semantic search

Cosine similarity over chunk embeddings retrieves the right context every time.

Private & temporary

24h, 7d, 30d expiry. Permanent accounts available via admin approval.

Shareable link

Unique Access ID + Model ID. Resume any time at /chat/AX92KD.

Dark / Light theme

Toggle anytime — full visibility parity in both themes.

Fast & light

AJAX everywhere. No heavy frameworks. Loads instantly on shared hosts.

Admin controls

Manage users, files, chunks, settings, requests, logs — all via AJAX.

How it works — RAG pipeline

PDF Upload → Extract → Clean → Chunk → Embed → Vector Search → Inject → AI answer.

Upload PDF

PDF / TXT / DOC / DOCX / MD / CSV up to 15MB.

Chunk + embed

Sentence-aware chunking with overlap, plus 256-dim semantic embeddings.

Retrieve top-K

Cosine similarity finds the most relevant context for your question.

Logfare answers

Logfare AI composes a natural, accurate reply grounded in your data.

Mini-lessons: how AI works

For Sarc Education Foundation students — quick, visual, and beginner-friendly.

1 What is RAG?

Retrieval-Augmented Generation means the AI first retrieves relevant text from your documents, then generates an answer using that context. This makes answers grounded and accurate.

  • Chunks — small searchable pieces of your PDFs.
  • Embeddings — vectors that capture meaning.
  • Top-K — the few most relevant chunks for your question.

2 Why hybrid AI?

SmartTrain AI is hybrid: greetings and general chat use Logfare AI directly, while data-grounded questions trigger the RAG pipeline. This keeps replies fast for small talk and accurate for documents.

3 How do embeddings work?

An embedding turns text into a list of numbers. Similar meanings produce similar numbers — so we can find related chunks just by comparing vectors with cosine similarity.

4 Privacy & expiry

By default, your workspace auto-deletes after the chosen expiry (24h / 7d / 30d). Need a permanent account? Submit a request — an admin will review it.

Ready to build your first AI?

It's free, private and takes less than a minute. No account needed.

Create my AI workspace