Hamza M.

I build RAG chatbots, AI agents and the data pipelines behind them, for teams that need a working system in production, not a demo. Fluent in English, French and Arabic. What I deliver • RAG & document AI: chatbots and search over your documents (PDFs, websites, legal and regulatory corpora), including French and Arabic content, with cited answers. • AI agents & automation: agents that qualify leads, answer customers and run workflows connected to your CRM (HubSpot, Keap) through n8n, Make or custom Python. • Data pipelines: ETL/ELT with Python, SQL and Spark on AWS and Azure, so your AI and dashboards run on clean, reliable data. Recent work • Insolvency firm in Québec (1,500+ clients/year): AI & automation architect, connecting their CRM and case-management tools and automating manual processes (worked in French). • Easelect: AI shopping assistant. Built the query-intelligence layer and product search engine (5★: "excellent candidate for AI and data integration"). • ATEM Analytics: Neo4j knowledge graph and branded RAG chatbot. • Arabic legal RAG & multi-agent system over a dense regulatory corpus, cutting document retrieval time by 60%+. • ContentFlow: 13-month data engineering engagement. Plus a production Spark job on AWS (S3, Athena). How I work I start by defining what "done" means (POC or production), then deliver in milestones with a demo at each step. You get clear updates, documented code and no surprises near the deadline. Stack: Python, FastAPI, LangChain, LangGraph, LlamaIndex, OpenAI, Claude · Pinecone, pgvector, Weaviate, Neo4j · n8n, Make · Spark, Airflow, PostgreSQL, Snowflake · AWS, Azure, Docker. Send me your use case and I'll reply with a concrete plan within 24 hours.