Projects
A collection of systems covering enterprise RAG, agentic workflows, computer vision, semantic search, recommendation engines, and applied NLP.
The projects that best show production scope, architecture depth, and measurable outcomes.
Architected AI-powered logistics platform using MCP agents & Strands framework with intelligent job assignment algorithms that automate delivery coordination and optimize driver efficiency.
100+ concurrent jobs, automated delivery coordination, dynamic route optimization
View case studyProduction-grade RAG pipeline with AWS Bedrock, OpenSearch, and multi-model LLM support featuring agentic tools and real-time streaming.
Multi-model LLM support, real-time streaming, agentic framework, production safety
View case studyEvent-driven AWS SageMaker pipeline for offshore inspection analysis using Claude Sonnet 4.5 and TwelveLabs for image/video processing.
30-40% cost reduction, 95%+ success rate, 100+ concurrent executions
View case studyEnterprise-grade multilingual chatbot for a major utility company using Claude 3 & Cohere v3 embeddings with AWS SageMaker integration for Arabic/English document processing.
Impact: Arabic/English support, enterprise-grade RAG, optimized response accuracy
Multi-agent AI system for real-time airline pricing with autonomous agents for price adjustments, market analysis & revenue forecasting across North America, Europe & Asia.
Impact: Significant revenue optimization, real-time price adjustments, global deployment
RAG-enhanced spell correction using Llama3-8B & Databricks with semantic retrieval, achieving 7.5% conversion rate increase and 40% token reduction.
Impact: 7.5% conversion increase, 40% token reduction, 18s latency improvement
Enterprise computer vision system using YOLO for real-time object detection of logbooks and starter packs, deployed globally across multiple operational environments.
Impact: Global deployment, real-time detection, enhanced operational efficiency
Built personalized recommendation system using entity recognition & knowledge graphs, significantly boosting user engagement and retention on news platform.
Impact: Increased user engagement, improved retention, real-time updates
Developed multilingual sentiment analysis model for large streaming company, improving viewer retention & platform engagement through enhanced personalization.
Impact: Improved retention, enhanced engagement, multilingual support
Designed internal HR system with NLP-based skill extraction & JD-CV matching, streamlining recruitment workflows and enhancing hiring decisions through intelligent automation.
Impact: Improved talent acquisition efficiency, automated candidate matching
Led development of image matching system enhancing retrieval accuracy by 30% using advanced vector-search algorithms, delivering high-performance image retrieval at scale.
Impact: 30% accuracy improvement, scalable architecture, reduced search times
Pioneered recommendation engine boosting user engagement by 43%, driving significant improvements in user retention and sales through intelligent personalization.
Impact: +43% engagement boost, improved retention, increased sales
Developed NLP model for semiconductor ticket management, cutting operational costs by 42% through streamlined automated ticket classification and routing.
Impact: 42% cost reduction, streamlined ticket management, automated routing
Spearheaded semantic search implementation enhancing product relevance by 32% and reducing query latency, significantly improving e-commerce search experience.
Impact: 32% relevance improvement, reduced query latency, better search UX