← All projects
ContactResume
Project

KaleshAI

A Hinglish evil-twin chatbot

Streamlit chatbot with a roast-style Hinglish persona, local Ollama or OpenAI-compatible LLMs, and document Q&A with source labels.

PythonStreamlitOllamaBM25LLM
scroll
Overview

What it is

KaleshAI is a mischievous Hinglish chatbot with a real LLM-backed chat flow and optional document Q&A. It is an evil-ish twin: quick roasts, light gaali, and it still answers the question.

Local or cloud LLM

Runs on a local Ollama model by default, or any OpenAI-compatible endpoint you configure.

Document Q&A

Upload TXT, MD or text-based PDF; text is chunked and retrieved with lexical BM25, then passed to the LLM.

Source labels

Answers from documents show which source chunks they came from.

Clean errors

Provider failures (HTTP errors, timeouts, bad replies) raise clear errors instead of crashing the chat.

Tested offline

20 unit tests with a mocked LLM client, so no paid model calls.

Flow

How it works

Pick a brain

Local Ollama, or an OpenAI-compatible base URL, model and key.

Optionally add documents

Upload files; they are chunked and indexed with BM25.

Chat

Matching chunks go to the LLM with the roast persona; answers carry source labels.

Source code only, run locally with Streamlit. Contains profanity by design.

Run it

git clone https://github.com/Prajit-22/KaleshAI.git cd KaleshAI pip install -r requirements.txt streamlit run app.py
Next projectIncident Triage →