t
tis_art

Taha Nawab

@tis_art

Reach me out to have better experience

Pakistan
English, Latin, Italian, German
About me
I am a Full Stack Developer with expertise in AI and Machine Learning. I have 3+ years of experience building applications and 1+ year working with AI and ML. I turn ideas into working software—full-stack apps on the MERN stack, REST APIs (Node, Python/FastAPI), and ML models deployed from notebook to live API. Stack: React, Node, MongoDB, PostgreSQL, MySQL, Firebase/Cloud Firestore, MediaPipe, BERT, HuggingFace, scikit-learn, n8n. I scope honestly, write clean maintainable code, and communicate in plain language. Message me with what you need to build—I'm happy to advise before you order.... Read more

Skills

t
tis_art
Taha Nawab
Offline • 

Work experience

Self_Employed

Full Stack Developer (MERN + ML models+ AI Automation)

Self Employed • Self-employed

Jun 2024 - Present • 2 yrs 1 mo

MediPredict — AI Medical Diagnosis Platform Built a full-stack disease-prediction app: trained a machine learning model (RandomForestClassifier) on 100+ symptom combinations, deployed it behind a FastAPI backend, and connected it to a React frontend. ~90% prediction accuracy, under 200ms response time. Includes JWT auth and a PostgreSQL database. Customer Support Ticket System with SLA Monitoring Enterprise-style ticketing platform built to handle 10,000+ tickets at once. Tickets auto-escalate to managers if unanswered for 4 hours, using a PostgreSQL trigger. Includes role-based agent assignment, a live analytics dashboard, and a full audit log for compliance. Stack: Node.js, Express, React, PostgreSQL. ToxiGlow — AI Wound Analysis Platform Built in a hackathon (BioNova Innovathon 2026): a computer vision tool that analyzes wound photos and returns tissue classification, severity score (0–100), and measurements — all in under 5 seconds. Includes role-based access for patients, clinicians, and admins, plus downloadable PDF reports. Stack: React, FastAPI, OpenCV, OpenAI API. VisionAI — OCR + Text-to-Speech Toolkit A Python tool that extracts text from images/scanned documents (Tesseract OCR) and reads it aloud automatically. Processes documents roughly 3x faster than manual transcription. Built with a modular structure so new AI features can be added easily. Warehouse Robot Pathfinding Simulation An AI simulation comparing pathfinding algorithms (A*, BFS, Genetic Algorithm) for robot navigation, with a live visual GUI showing each decision in real time. A* found the optimal path 95% of the time. Lead Capture & CRM Automation (n8n) Built an n8n workflow that automatically captures form submissions, syncs them to a CRM/spreadsheet, and sends a notification — removing manual data entry from the lead-intake process.