Mathieu P.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

Most automations work in a demo and break the moment they hit real data. I build the ones that hold up. 𝗜’𝗺 𝗮 𝘀𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗱𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿 who also runs operations for a distribution client, so I’ve been on both sides of a system that has to work daily. That company does eight figures across 3,000+ retail accounts. The order flows, invoicing logic, EDI feeds, and system integrations I build there run on messy real-world data with actual money moving through them. When something breaks, it’s my problem. That shapes how I build for clients. 𝗧𝗵𝗲 𝗴𝗮𝗽 𝗜 𝘄𝗼𝗿𝗸 𝗶𝗻. Modern tools and AI get you most of the way there, and that’s genuinely useful. What they don’t handle is the part where your customer names exist in three different formats, the API you need has undocumented rate limits, the legacy system has no clean export, and the workflow fails silently on the one edge case that matters. That’s where projects stall, and it’s the work I’m built for. 𝗔𝗜 𝗶𝘀 𝗼𝗻𝗹𝘆 𝗮𝘀 𝗴𝗼𝗼𝗱 𝗮𝘀 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮 𝘂𝗻𝗱𝗲𝗿𝗻𝗲𝗮𝘁𝗵 𝗶𝘁. Standing up a chatbot is easy now. Connecting a call tracking tool or a support agent takes an afternoon. But if the data feeding it is duplicated, inconsistent, or scattered across systems that disagree with each other, the output is confidently wrong and you’ve automated a problem instead of solving it. Getting a company’s data into a state where AI can actually use it is the part nobody demos, and it’s most of the real work. I structure the data first, then build on top of it. 𝗜 𝗱𝗶𝗮𝗴𝗻𝗼𝘀𝗲 𝗯𝗲𝗳𝗼𝗿𝗲 𝗜 𝗯𝘂𝗶𝗹𝗱. A lot of automation requests are a fix for the wrong problem. Before I write anything I want to understand what the process costs you now and what changes when it’s gone. Sometimes the honest answer is a smaller project than you expected. I’d rather tell you that than bill you for it. 𝗡𝗼 𝗰𝗲𝗶𝗹𝗶𝗻𝗴. Most automation specialists stop where their platform stops. My background is full-stack development, so when Make, Zapier, n8n, or Airtable hits its limit, I write the code to go past it. Custom API endpoints, scripting, middleware, proper error handling and logging so you know when something fails instead of finding out from a customer. 𝗪𝗵𝗼 𝘁𝗵𝗶𝘀 𝘄𝗼𝗿𝗸𝘀 𝗳𝗼𝗿. Businesses with enough volume that the math is obvious: processes designed when the company was half its current size, data living in four places, a team doing by hand what the system should do. If I can’t see a project paying for itself inside a year, I’ll say so before you spend the money. 𝗪𝗵𝗮𝘁 𝗜 𝗯𝘂𝗶𝗹𝗱: ✓ Connect CRM, website, accounting, and inventory systems so data stops living in four places ✓ Clean up and restructure business data so AI and reporting actually work on it ✓ Quoting, invoicing, and approval workflows with proper validation ✓ Airtable and internal database systems that replace spreadsheet workarounds ✓ AI integrated where it produces measurable return, not demos ✓ Custom API development when off-the-shelf connectors fall short ✓ Repair and rebuild of automations that work most of the time and fail on edge cases ✓ Platform migrations, including Zapier to Make or n8n when task billing gets expensive ✓ Custom scripts covering a broad range of automation, for example lead generation from Google Search Results (SERP) 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀 𝗜 𝘄𝗼𝗿𝗸 𝘄𝗶𝘁𝗵: Systems Integration, Data Architecture, Data Modeling, Data Pipeline, Data Quality, Data Cleaning, Data Normalization, Deduplication, Record Matching, Single Source of Truth, Data Migration, Data Transformation, ETL, Database Design, Business Process Automation, Operations Automation, Workflow Automation, Automation, Automation Audit, Error Handling, API Integration, API Development, REST API, Webhooks, Middleware, CRM Integration, ERP Integration, EDI, n8n, Make (Integromat), Zapier, Airtable Scripting, HubSpot, Salesforce, Pipedrive, QuickBooks, Stripe, Shopify, Google Sheets, Slack Integration, Webflow, AI Integration, LLM Integration, AI Agent, AI Chatbot, RAG, LangChain, Knowledge Base, OpenAI, Claude, Document Processing, Data Extraction, Full Stack Development, Backend Development, JavaScript, Python, Node.js, Invoice Automation, Order Management, Inventory Systems, SERP Data (Maps & Reviews), Local Business, Data Enrichment, Prospect List, Google Tag Manager, Analytics. 
𝗤𝘂𝗶𝗰𝗸 𝗳𝗶𝘅𝗲𝘀 ✓ Make: scenario disabled, incomplete executions, temporarily paused, missing error handlers, operations usage ✓ Zapier: Zap turned off, held Zap runs, Throttled by Zapier, Zapier has blocked this task, task usage ✓ n8n: webhook not triggering, expression and paired item errors, execution timeout, expired credentials, continue on fail ✓ Data not syncing, integration stopped syncing, records not updating, duplicate records, API rate limits ✓ Invoice parsing, email order intake, EDI, Stripe invoices ✓ Automation stopped working, rollback, sub-workflows, binary data, queue mode, self-hosted