Python/FastAPI AI Backends
RAG APIs, document ingestion, embeddings, vector stores, source-grounded answers, namespaces, auth, queues, and service integrations.
Python | React | Laravel/PHP | AI Systems
I build AI products with Python/FastAPI, React, Laravel/PHP, vector search, LLM agents and automation workflows. With 5+ years in software delivery, my strength is turning fuzzy AI product ideas into working systems with clear scope, review, testing and delivery.
What I Do
I work where AI features become actual software: FastAPI services, React dashboards, ingestion pipelines, vector search, APIs, queues, admin surfaces, integrations, and the client feedback loop around them.
RAG APIs, document ingestion, embeddings, vector stores, source-grounded answers, namespaces, auth, queues, and service integrations.
Dashboards, admin surfaces, chat workflows, API-key screens, usage views, and product flows that make AI systems usable.
Laravel AI packages, PHP production apps, chatbot SaaS workflows, WhatsApp integrations, CRM systems, and business automation.
Selected Work
Some production code stays private. I document those projects through sanitized case studies, landing pages, and screenshots while keeping source code and business logic protected.
Advanced RAG platform built around FastAPI and React for secure document intelligence with namespaces, ingestion, grounded chat, external API access, memory, billing, usage, and admin workflows.
Python benchmark runner delivery for LLM optimization workflows: CLI, artifacts, resume and timeout behavior, Gradio UI, MLflow, TensorBoard, and validation notebooks.
Laravel-native RAG and AI agent toolkit with document ingestion, vector search, streaming chat, multi-provider AI support, tenant scoping, and health checks.
Multi-tenant Laravel/PHP chatbot SaaS powered by LarAI Kit. The public showcase covers chatbot setup, knowledge upload, chatbot detail, WhatsApp integration, API tokens, widget embed, and admin flows.
Master's thesis on a 50,000-review Roman Urdu dataset and hybrid 1D+2D CNN model for low-resource sentiment classification.
Stack
Python, FastAPI, LangChain, LangGraph, OpenAI, Anthropic, Gemini, Pinecone, Qdrant, FAISS, pgvector, MLflow, TensorBoard
React, TypeScript, Vite, Redux, dashboards, admin panels, chat interfaces, API-key screens, product workflows
Laravel, PHP, CodeIgniter, package design, REST APIs, webhooks, JWT, queues, CRM systems, WhatsApp integrations
n8n, Google Sheets API, WhatsApp Business API, Twilio, Stripe, Salesforce, Tableau, ClickUp, Acumatica
Experience
Lead AI engineering delivery across Python/FastAPI RAG systems, React dashboards, voice agents, LangGraph/n8n automations, WhatsApp Business API integrations, CRM workflows, and intern mentoring.
Delivered React frontend work, CodeIgniter/PHP backends, CRM customization, Dolby.io integrations, WordPress/PHP work, and production API workflows.
Built admin panels, dashboards, REST APIs, jQuery/AJAX workflows, and automated data processing systems.
Education And Research
MS Data Science - FAST NUCES, Lahore
BS Computer Science - Government College University, Lahore
Research focus: Roman Urdu sentiment analysis using machine translation and a hybrid 1D+2D CNN approach.