Career
Experience
Roles aligned with shipped systems — ERP GenAI at Royal Soft, solo FYP recommender, and independent AI products.
Aug 2025 — Present · Lahore, Pakistan · on-site
AI & ML Developer
Royal Soft · full time
Shipping production GenAI systems for enterprise ERP: multi-agent chatbots, codebase-aware PHP generation, n8n SDLC automation, and meeting-to-tasks tooling.
- Built and operate the Intelligent ERP Chatbot — multi-tenant LangChain agents, safe SQL tools, forecasting, Plotly KPIs, and EN/Urdu responses against live MySQL.
- Delivered the Royal ERP AI Code Generator — ZIP ingest, convention parsing, retrieval, orchestrated PHP/SQL/JS generation, validators, and packaging.
- Shipped the SDLC Automation Bot (Next.js + n8n + Recall.ai + Claude/TMS) turning meetings into structured requirements and assigned tasks.
- Built the Codebase-Aware Prompt Engine (n8n + MySQL) to ground developer prompts in real ERP file excerpts and batch history.
- Developed the Meeting Recorder / join-bot stack (Django, Celery, Playwright, Faster-Whisper) for automated capture and transcription.
- 500+ daily natural-language ERP queries via multi-agent orchestration
- ~40% fewer manual lookups; ~35% faster query resolution
- Code generation grounded on 1,400+ ERP files — 70–90% convention similarity; 60%+ less manual effort
- 60%+ faster project kick-off through SDLC meeting → task automation
PythonDjangoLangChainOpenAIMySQLn8nNext.jsChromaDBPlotlyAnthropicPostgreSQLCeleryPlaywright
Aug 2024 — Jun 2025 · Bahawalpur, Pakistan · on-site
AI Engineer — Final Year Project (Solo)
Islamia University of Bahawalpur · self employed
Solo Final Year Project: full Django e-commerce platform with hybrid collaborative + content-based product recommendations (85%+ validation precision).
- Solo-built full e-commerce storefront and admin (Django, Bootstrap, media pipeline).
- Implemented RecommendationEngine with collaborative filtering, TF-IDF content similarity, hybrid merge, and persistence.
- Engineered interaction weights (view/search/like/cart/purchase) and generate_recommendations management command.
- Validated recommendations at 85%+ precision; deployed with Django + MySQL (PythonAnywhere path).
- 85%+ recommendation precision on held-out validation
- End-to-end solo FYP delivery over ~10 months
PythonDjangoScikit-learnPandasNumPyMySQLBootstrap
Jan 2024 — Present · Lahore, Pakistan · remote
AI & Full-Stack Builder
Independent / Personal Products · self employed
Personal production-style builds: Smart AI Resume Analyzer, MediCare ML predictions, Superstore POS/ERP, CampusGuide AI, and frontend AI utilities.
- Smart AI Resume Analyzer — ATS scoring, Groq feedback, job match, PDF/DOCX builder (Django REST + React).
- MediCare — diabetes and depression risk ML pipelines with validation and structured prediction APIs.
- Superstore Management System — POS, inventory, accounting, CRM/HR (Django REST + React/Vite).
- CampusGuide AI — structured college data + hybrid RAG assistant.
- Multiple end-to-end products published on the portfolio with full case studies
PythonDjangoDjango REST FrameworkReactViteScikit-learnGroqLangChainPostgreSQL
May 2023 — Sep 2023 · Bahawalpur, Pakistan · on-site
Python & ML Intern
Code Lab Tech School & IT Solutions · internship
Foundational internship in data analysis, classical ML, visualization, and a Django + MySQL capstone — base for later production GenAI work.
- Completed hands-on training in Pandas, NumPy, Scikit-learn, Matplotlib, and Seaborn across multiple data projects.
- Delivered a Django + MySQL web application as the internship capstone.
- Practiced feature engineering, model evaluation basics, and clean experiment notebooks.
- 3 functional data/ML projects completed during internship
- Django + MySQL capstone delivered
PythonDjangoScikit-learnPandasNumPyMySQL
Education
Academic background