Ahsan U.

𝐘𝐨𝐮𝐫 𝐭𝐞𝐚𝐦𝐬 𝐚𝐫𝐞 𝐛𝐮𝐫𝐧𝐢𝐧𝐠 𝟏𝟎𝟎 𝐡𝐨𝐮𝐫𝐬 𝐚 𝐰𝐞𝐞𝐤 𝐨𝐧 𝐫𝐞𝐩𝐨𝐫𝐭𝐢𝐧𝐠/𝐫𝐞𝐩𝐢𝐭𝐚𝐭𝐢𝐯𝐞 𝐩𝐫𝐨𝐜𝐞𝐬𝐬 𝐚𝐧𝐝 𝐢𝐧𝐭𝐚𝐤𝐞 𝐰𝐨𝐫𝐤 𝐀𝐈 𝐜𝐚𝐧 𝐚𝐥𝐫𝐞𝐚𝐝𝐲 𝐝𝐨 𝐞𝐧𝐝 𝐭𝐨 𝐞𝐧𝐝. 𝐈 𝐛𝐮𝐢𝐥𝐝 𝐭𝐡𝐞 𝐬𝐲𝐬𝐭𝐞𝐦𝐬 𝐭𝐡𝐚𝐭 𝐞𝐥𝐢𝐦𝐢𝐧𝐚𝐭𝐞 𝐢𝐭. Production AI systems shipped across insurance, property management, e-commerce, manufacturing, Judiciary & Law and enterprise data operations. Not prototypes. Real systems running in production, monitored, and built to survive contact with messy real-world data. 𝐑𝐞𝐜𝐞𝐧𝐭 𝐏𝐫𝐨𝐣𝐞𝐜𝐭𝐬: ✓ Adjustor Voice AI Agent for Real-Time Insurance Claim Intake. Replaces hold music and repeated forms with a real-time voice agent that understands the claim as it's spoken. [Voice AI, LLM, Real-Time Intake through Google ADK,Lite LLM , Agentic AI , AWS] ✓ AI Receptionist + RAG Knowledge Base (O-Lac Chalet, LIVE). Full retrieval pipeline, sitemap crawl, chunking, OpenAI embeddings, Pinecone vector store, feeding a LangChain agent that answers guest questions and checks live calendar and property availability. Deployed in production handling both voice (Retell AI) and chat. [LangChain, RAG, Pinecone, Voice AI] ✓ AI Text-to-SQL Engine, LangGraph Agent with Semantic Schema Search. Agentic SQL generation that retrieves only the relevant tables per query instead of dumping the whole schema at the model. [LangGraph, RAG, ChromaDB] ✓ Enterprise Sitemap Sync Platform (Subway). Real-time crawling and sync across Subway's production web properties, replacing manual sitemap tracking. [Node.js, Notion API, Vercel] ✓ Real-Time Reporting Automation, Power BI + n8n. Eliminated 20+ hours of weekly manual reporting by wiring live dashboards directly to automated workflows. [Power BI, n8n, Workflow Automation] ✓ ETL Pipeline & Data Warehouse Architecture. Processing 2.4B+ rows daily from 18 sources into Snowflake and Delta Lake. [Airflow, dbt, Snowflake] ✓ AI Text-to-SQL Engine, LangGraph Agent with Semantic Schema Search. Agentic SQL generation that retrieves only the relevant tables per query instead of dumping the whole schema at the model. [LangGraph, RAG, ChromaDB] ✓ Enterprise Document Intelligence App. Intelligent document parsing and classification using LLMs and RAG, rated elite by the client. [LLM, RAG, Document AI] ✓ Property Management Voice AI with Automated Maintenance Ticketing. Handles resident questions and logs maintenance tickets without a human on the line. [Voice AI, Ticketing Automation] 𝐖𝐡𝐚𝐭 𝐈 𝐛𝐮𝐢𝐥𝐝 𝐟𝐨𝐫 𝐜𝐥𝐢𝐞𝐧𝐭𝐬: ✓ AI agents and voice agents for intake, support, and operations, built for real production reliability ✓ RAG chatbots and document intelligence using OpenAI, Claude, LangChain/LangGraph, and vector search ✓ Power BI and data warehouse pipelines (Snowflake, BigQuery, Airflow, dbt) wired to automated reporting ✓ Custom API integrations and CRM/workflow automation (n8n, Make) ✓ Clean, governed data underneath every AI system, not a fragile demo sitting on top of nothing 𝐖𝐡𝐲 𝐜𝐥𝐢𝐞𝐧𝐭𝐬 𝐡𝐢𝐫𝐞 𝐦𝐞: * Every system ships with error handling and monitoring your team can actually see, not a script that breaks in a week * I started in data engineering and Power BI, so the AI layer sits on real, trustworthy data, not guesses * End-to-end ownership from the first conversation to the system running in production 𝐆𝐨𝐨𝐝 𝐟𝐢𝐭: if you have a real workflow burning real hours every week and want it gone for good. Not a fit if price is the only thing that matters, or you need results in under two weeks. If you're automating a manual process or connecting AI to your real tools, send me a short description of the workflow and I'll map the safest architecture with you before we build anything.