arslanahmed.me · LinkedIn · Fiverr · Email
I build production software across mobile, backend, and applied AI, and I care more about shipping AI into real products than running experiments in notebooks. AgriGuard started with a client requirement, ran through six months of EfficientNetB3 training on crop disease data, and ended with an ESP32 controlling an irrigation pump from a React Native app. That's the kind of problem I find interesting.
Freelancing internationally since my second year of university. 5.0 on Fiverr. Currently shipping SpendWise.
Built for AG Leaders. A single disease classifier fails because it conflates diseases across crops: a tomato lesion and a mango lesion look similar enough to confuse the model. The fix: a crop gate first. EfficientNetB3 identifies the crop (11 classes, 99.15% accuracy), rejects unknown inputs, then a second EfficientNetB3 classifies disease within that confirmed crop (76 classes, 94.86% accuracy, 0.94 macro F1). Dual ESP32 architecture: CAM module handles image capture, WROOM handles pump relay and soil moisture, separated because running GPIO alongside camera and WiFi on a single ESP32-CAM causes ADC failure.
React Native Node.js TensorFlow Flask EfficientNetB3 ESP32 HuggingFace
SpendWise · Live · Google Play
AI finance tracker built solo, live on Android. The financial advisor is RAG-grounded against each user's transaction history, budgets, and wallets before generating responses, preventing generic or fabricated advice. Gemini Vision extracts structured transaction data from receipts, reducing manual entry to a single image. Native Android home screen widget built in Kotlin with Jetpack Glance so users can log transactions without opening the app. AI features routed through a Next.js backend API to keep keys off the client. Jest test coverage before release.
React Native Kotlin TypeScript Expo Firebase Gemini Groq Next.js RAG Jest
AI recruitment platform. Backed by PostgreSQL on Supabase to handle the concurrent read/write patterns an applicant tracking system generates: multiple recruiters reviewing, scoring, and updating candidates simultaneously. LangChain + Groq handle automated screening interviews with instant scoring. Stripe subscription tiers for premium features.
Next.js Django REST PostgreSQL LangChain Groq Stripe Supabase
ReelTime · Kaffeine · Movies Recommending System · Employee Management System
Languages JavaScript · TypeScript · Python · Kotlin · C# · C++
Mobile React Native · Expo · Android SDK
Frontend React · Next.js · Tailwind CSS
Backend Node.js · Express · Django REST · Flask
Databases MongoDB · PostgreSQL · Firebase · MySQL
AI/ML TensorFlow · PyTorch · Keras · LangChain · Scikit-learn · Pandas · NumPy
Tools Git · Linux · Postman · Android Studio
Python Developer — AI/ML · Smart Algorithm (AI Soft) · Dec 2024 – Mar 2025
Designed, trained, and deployed seven ML models across classification, regression, and clustering. Integrated trained models into production Django applications end-to-end.
Freelance Developer · Fiverr · Jan 2024 – Present
5.0/5.0. Mobile apps, REST APIs, and full-stack systems for international clients.