Hire Expert AI Engineers

OpenWriterAI — Lahore, Pakistan

• Architected and launched OpenWriterAI, a production-grade subscription AI SaaS platform with a decoupled Nest.js backend and reactive Next.js frontend
• Built a secure multi-tenant ecosystem with LangChain, OpenAI APIs, and a Supabase vector database (RAG) powering instant resume generation and text humanization
• Developed a cross-platform companion browser extension interfacing with core APIs over secure WebSockets for low-latency on-page writing assistance
• Integrated multi-tenant authentication and Lemon Squeezy webhook gateways, automating subscription lifecycle events and recurring billing pipelines

U.S. AI Venture (Pegasus Project) — Remote, U.S. Contract

• Architected and deployed an enterprise-grade local-first multimodal AI chatbot platform on Ollama and FastAPI, delivering data-private alternatives to cloud solutions
• Engineered self-correcting agent execution loops with LangChain and LangGraph, integrating autonomous browser automation to research and verify live web data
• Built a high-throughput async WebSocket backend for real-time voice streaming and live video chunk analysis, achieving sub-3-second streaming response
• Designed a node-based admin workflow interface in Next.js enabling real-time multi-model routing, hot-swapping, and automated failover
• Reduced VRAM footprints by 40% through model quantization and tensor parallelism while maintaining high token-generation throughput

Private Equity Professional — Remote, U.S. Contract

• Architected an enterprise-grade RAG pipeline ingesting, structuring, and indexing 100,000+ private newsletter articles for contextual LLM search
• Built an automated n8n data ingestion pipeline handling HTML extraction, text chunking, and vector database syncing to keep the chatbot current
• Developed a custom WordPress Single Sign-On integration securing authentication for active portal subscribers
• Engineered the full stack in Python (FastAPI/Django) and React/Next.js, optimising API retrieval loops for low-latency response streaming
• Slashed GPT API costs and token consumption by 40% via prompt caching, semantic token pruning, and vector meta-filtering

IkamGroup — Remote, U.K. Contract

• Architected FacelessLab, an autonomous video generation and publishing platform automating the full YouTube content lifecycle from a single reference link
• Built data collection pipelines scraping YouTube metadata and transcripts, feeding fine-tuned GPT models to generate tailored scripts
• Integrated Imagen 4 and Nano Banana image pipelines for programmatic head-swapping and canvas-drawing on high-conversion thumbnails
• Engineered async workflows streaming scripts to the Revid.ai API and auto-publishing finished media via the YouTube Data API
• Implemented Redis/Celery job queues managing concurrent renders, preventing API timeouts and accelerating compilation speeds by 35%

Rise Marketing Engineer — Remote, U.S. Contract

• Engineered and deployed a production-ready interactive real estate management platform from architecture to live deployment in Next.js and NestJS
• Built an interactive plot mapping and inventory dashboard broadcasting real-time availability across embedded client views
• Implemented secure multi-tenant infrastructure on Supabase with role-based access control and scalable TypeScript REST APIs
• Managed production infrastructure, deployment, and monitoring on Render, optimising client-side rendering for third-party embeds
• Boosted map embed loading speeds by 45% through PostGIS spatial index optimisation and aggressive Redis caching of plot geometry

U.S. E-Commerce Venture (Attronaut Portal) — Remote, U.S. Contract

• Engineered Attronaut, an enterprise-grade multi-tenant e-commerce analytics platform unifying real-time storefront revenue and ad performance data
• Configured secure OAuth 2.0 flows and webhook gateways ingesting live marketing, spend, and conversion metrics from Shopify and Google Ads
• Built high-throughput Django backends with asynchronous background workers running scheduled data-gathering loops, eliminating manual reporting overhead
• Developed a responsive Next.js and Tailwind CSS dashboard with sub-second interaction latency and domain-verified email gateways

Programmers Venture — Lahore, Pakistan

• Deployed scalable Python (Django) and JavaScript enterprise backends integrating custom ML models that boosted data-forecasting accuracy by 25%
• Architected real-time computer vision (CNN) pipelines streamlining live object detection with zero frame degradation
• Built context-aware NLP modules improving automated real-time customer sentiment classification
• Engineered secure Django REST middleware eliminating common API vulnerabilities and safeguarding multi-tenant data access points
• Migrated a monolithic system to an optimised microservices-driven architecture with cross-functional Agile teams