Connor O.

AI engineer building production AI agents, RAG pipelines and LLM apps for teams that need more than a demo. Architected a ~100-route FastAPI legal AI platform and a 36,000-line WMS/ERP, and grew my own recommerce business 87% year over year. If your AI works in a demo but breaks on real documents, real users or real volume, that is the gap I close. 𝗪𝗛𝗔𝗧 𝗜 𝗕𝗨𝗜𝗟𝗗 • 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 𝗮𝗻𝗱 𝗟𝗟𝗠 𝗮𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀: agentic workflows, AI copilots and decision-support tools on OpenAI and Anthropic models, with structured outputs, tool use and human review where it matters. • 𝗥𝗔𝗚 𝗮𝗻𝗱 𝗸𝗻𝗼𝘄𝗹𝗲𝗱𝗴𝗲 𝘀𝘆𝘀𝘁𝗲𝗺𝘀: document intelligence, semantic search, vector embeddings and Neo4j knowledge graphs, so every answer is grounded and traceable to its source. • 𝗔𝗜 𝗮𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗶𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻𝘀: business process automation, CRM and API integrations, data enrichment and workflow orchestration that runs without babysitting. 𝗥𝗘𝗦𝗨𝗟𝗧𝗦 • 𝗟𝗲𝗴𝗮𝗹 𝗔𝗜 (Juricratic, building now): architected a five-package Python/TypeScript monorepo with nearly 100 FastAPI routes, semantic retrieval, a Neo4j knowledge graph, Bayesian risk modelling and human-review controls. • 𝗥𝗲𝘃𝗲𝗿𝘀𝗲 𝗹𝗼𝗴𝗶𝘀𝘁𝗶𝗰𝘀 (QuickWMS): designed and built a 54-module, ~36,000-line WMS/ERP on my own, covering receiving, barcode scanning, grading, refurbishment, marketplace listing, fulfilment and analytics. • 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝘃𝗶𝘀𝗶𝗼𝗻 (QuickVisionz): an edge AI workflow on Raspberry Pi 5 and a Hailo NPU with FastAPI and Next.js, demonstrated live on a conveyor at ASD Market Week, Las Vegas. • 𝗥𝗲𝗮𝗹 𝗲𝘀𝘁𝗮𝘁𝗲 (AIWholesail): a lead intelligence platform with seller scoring, data enrichment, outreach automation and CRM workflows. • 𝗥𝗲𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗲 (Upscaled): founded it and grew it 87% year over year through workflow automation, pricing and inventory discipline. • 𝗖𝗼𝗿𝗽𝗼𝗿𝗮𝘁𝗲 𝗵𝗼𝘂𝘀𝗶𝗻𝗴 (Sleepover): co-founded it and ran technology and operations across 12 states, held 92% occupancy through the pandemic, and exited at 24. 𝗛𝗢𝗪 𝗜 𝗪𝗢𝗥𝗞 1. Start from the outcome: what the system must do, for whom, and how we will know it works. 2. A written plan with milestones, risks and trade-offs before the heavy build. 3. Working increments you can test, evaluated on your real data rather than hand-picked examples. 4. Production basics from day one: logging, error handling, security, tests and documentation, so your team can own it after handoff. You get clear progress updates and early warning on risks, never surprises. 𝗜 𝗔𝗟𝗦𝗢 𝗛𝗘𝗟𝗣 𝗪𝗜𝗧𝗛 Full-stack SaaS and backend engineering, computer vision and edge AI, data pipelines and ETL, and WMS/ERP, inventory and e-commerce systems. 𝗧𝗢𝗢𝗟𝗦 𝗔𝗡𝗗 𝗧𝗘𝗖𝗛 ▪️ 𝗔𝗜 & 𝗟𝗟𝗠: OpenAI, Anthropic, AI Agents, LLM Applications, RAG, Prompt Engineering ▪️ 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲𝘀 & 𝗕𝗮𝗰𝗸𝗲𝗻𝗱: Python, TypeScript, JavaScript, FastAPI, REST APIs ▪️ 𝗙𝘂𝗹𝗹-𝗦𝘁𝗮𝗰𝗸 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁: Next.js, React, SaaS, Web Applications, Backend Architecture ▪️ 𝗗𝗮𝘁𝗮 & 𝗞𝗻𝗼𝘄𝗹𝗲𝗱𝗴𝗲 𝗦𝘆𝘀𝘁𝗲𝗺𝘀: Knowledge Graphs, Neo4j, Vector Embeddings, Semantic Search, ETL Pipelines, Data Lakes ▪️ 𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 & 𝗩𝗶𝘀𝗶𝗼𝗻: Machine Learning, Computer Vision, Image Processing, Model Evaluation, Edge AI ▪️ 𝗔𝗜 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻: Workflow Automation, API Integrations, CRM Automation, Agentic Workflows, Human-in-the-Loop Systems ▪️ 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 & 𝗜𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲: SQL, MongoDB, Vector Databases, Git, GitHub, CI/CD ▪️ 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗦𝘆𝘀𝘁𝗲𝗺𝘀: WMS/ERP, Inventory Management, E-commerce, Marketplace Integrations, Operational Analytics Send me your workflow, your data and what "working" looks like for you, and I will reply within one business day with how I would build it.