0%

19.0330° N,73.0297° E
mohammadjaan

/ AI Engineering

/ Machine Learning

/ Software Development

basedinnavi mumbai

I write code and train models. Some projects turn out great, others teach me what not to do next time. Either way, I'm building.

a b o u t  m e

1/5

about me

intro/1

about me

Hello!

I'm Mohammad Jaan

my journey →

I recently completed my Computer Engineering degree from K. J. Somaiya Institute of Technology. I spend most of my time working with AI, machine learning, and full-stack development, building things that solve real problems and exploring what's possible with technology.

r e c e n t  w o r k s

agent.py
import sys
import asyncio

# CRITICAL: Set event loop policy FIRST before any async imports
if sys.platform == 'win32':
    import nest_asyncio
    asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())
    nest_asyncio.apply()  # Allow nested event loops

from contextlib import asynccontextmanager
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import JSONResponse
import uvicorn

Autonomous Job Application Agent

AI-Powered Job Search Automation

Architected an autonomous AI agent that orchestrates the scraping, analysis, and application process for 200+ daily job listings across 6+ Job Listing platforms. Built a real-time analytics dashboard using Next.js 14 to track application status, reducing manual job search overhead by 80% (approx. 15 hours/week saved).

DebateEngine.tsx
// DebateEngine.tsx
import { useEffect, useState } from 'react';
import { useDebateStore } from '@/store/debateStore';
import { GoogleGenerativeAI } from '@google/generative-ai';

interface Persona {
  name: string;
  philosophy: 'utilitarian' | 'deontological' | 'virtue' | 'care' | 'contractarian';
  systemPrompt: string;
}

const personas: Persona[] = [
  { name: 'Marcus', philosophy: 'utilitarian', systemPrompt: '...' },
  { name: 'Immanuel', philosophy: 'deontological', systemPrompt: '...' },
];

Ethical AI Debate Simulator

Multi-Agent Philosophical Debate Platform

Engineered a multi-agent debate platform that simulates complex ethical dilemmas, utilizing Google Gemini API to power 5 distinct philosophical personas (Utilitarian, Deontological) with consistent reasoning capabilities. Developed a high-performance frontend with React 18 and Zustand to handle real-time stream updates.

plantClassification.py
# plantClassification.py
from ultralytics import YOLO
import cv2
import numpy as np

class PlantClassifier:
    def __init__(self, model_path='herbvision_best.pt'):
        self.model = YOLO(model_path)
        
    def classify_plant(self, image_path):
        results = self.model(image_path)
        
        for result in results:
            boxes = result.boxes
            for box in boxes:

HerbVision

Ayurvedic Plant Identifier

Deep Learning Mobile App for plant recognition achieving 93% mAP using YOLOv8 for real-time object detection on 200+ Ayurvedic medicinal plants across 5000+ curated images. Fine-tuned and quantized the model from 120MB to 40MB (67% reduction) for mobile deployment while maintaining 91% accuracy.

e x p e r i e n c e

July 2025 – October 2025

Andheri E, Mumbai

TickleRight Eduventures Pvt. Ltd.

Software Developer Intern

  • →Engineered 'TickleGram,' a centralized customer engagement platform, by integrating Instagram/Facebook Graph APIs and WebSockets, enabling real-time synchronization of 5,000+ daily messages across 50+ client accounts.
  • →Reduced support team response latency by 40% by architecting a unified chat interface with automated ticket assignment and status tracking features.
  • →Designed a secure video streaming infrastructure for the B2B Teacher Training platform, implementing AWS CloudFront Signed URLs and HLS chunking to eliminate unauthorized content downloads and reduce bandwidth costs by 15%.

January 2025 – June 2025

Remote

ExcelR Solutions

Full-stack Development Intern

  • →Developed the 'Mumbai Tourist Portal' using React, Node.js and MongoDB, implementing JWT-based authentication.
  • →Optimized frontend rendering performance by 30% through the implementation of React.memo and lazy loading strategies for image-heavy attraction galleries.
  • →Integrated Google Maps API to provide real-time routing and distance calculation, enhancing the itinerary planning experience for end-users.

December 2023 – February 2024

Remote

Claidroid Technologies Pvt. Ltd.

Machine Learning Intern

  • →Developed a high-accuracy botanical classification system using YOLOv8, implementing advanced transfer learning techniques to handle class imbalance across 200+ Ayurvedic plant species.
  • →Architected a robust data ingestion workflow, collecting and annotating over 2000+ images to create a proprietary dataset, validating ground truth labels to ensure 89% model accuracy.
  • →Optimized the deep learning model for mobile devices using TensorFlow Lite (TFLite), reducing model size by 60% to ensure low-latency inference.

Education

B.Tech in Computer Engineering

K. J. Somaiya Institute of Technology

CGPA: 9.102021 - 2025

Honours in Artificial Intelligence and Machine Learning

K. J. Somaiya Institute of Technology

Grade: 81%2023 - 2025

s k i l l s  &  t o o l s

Python

Expert

JavaScript

Expert

TypeScript

Advanced

C/C++

Advanced

SQL

Intermediate

HTML/CSS

Expert

React

Advanced

Next.js

Advanced

Node.js

Intermediate

PHP

Intermediate

MongoDB

Intermediate

PostgreSQL

Intermediate

PyTorch

Intermediate

TensorFlow

Intermediate

Vertex AI

Intermediate

GCP

Intermediate

FastAPI

Advanced

Git

Expert

Docker

Beginner

AWS

Intermediate

AI & ML

Generative AI, Machine Learning, NLP, Reinforcement Learning, Neural Networks, Game Theory

Full Stack

MERN Stack, RESTful APIs, Database Design, Server Deployment

Tools & Practices

Git/GitHub, Docker, CI/CD, Agile Methodologies, Postman

Let's start the conversation

G r e a t  p r o j e c t s

start with

g r e a t  c o l l a b o r a t i o n