Monica T.

✅ 135 real invoices audited before an automation went live, zero false alarms since ✅ Built a lead-discovery system that found 45,000+ companies with zero overlap against a paid vendor list ☑️ Backend developer who writes the code behind HubSpot, ActiveCampaign and Zapier, not just the no-code layer on top 🔍 HubSpot · ActiveCampaign · Zapier · Python · Supabase · Google Cloud Platform 📞 Invite me to your project or send a direct message to start a conversation. Hi, I'm Monica! I'm a backend developer who builds the Python code and API logic that make a company's HubSpot, ActiveCampaign and Zapier systems actually work together. I work full time with Growth Geyser, a RevOps and CRM agency, building and fixing the backend systems behind our clients' CRMs and integrations, for companies in telecom software and lab equipment supply. Recent work, with the numbers: ✅ Built a dunning-to-task automation that creates HubSpot tasks automatically when a customer's payment collection cycle fails, validated against 135 real invoices across 9 accounts before going live, then verified against a real case in production. ✅ Built a Python lead-discovery system from scratch using public DNS records instead of a paid list provider: found 45,325 companies with zero overlap against an existing paid vendor list of similar size, still running today. ✅ Found and fixed a $175,118.57 revenue-attribution gap across 70 companies by repointing a payment webhook to the correct field, verified against real payment records before and after. ✅ Recovered a production analytics system the same day a site migration broke it: rebuilt the tracking, fixed four report bugs, and had live data flowing again within hours. ✅ Built centralized error logging for a client's automation bot: a database table, automatic inserts every cycle, and a daily routine that only alerts when something actually needs attention. Before Growth Geyser, I spent two years at Continental Automotive writing software for systems where a bug could mean a safety failure. That is where the habits below come from. How I work: ✅ Nothing goes live until I have checked it against real data. The dunning automation ran against 14 months of real invoice history before I trusted it. ✅ Diagnose before I touch anything. When a client asked whether automation was misrouting support tickets, I proved it was not, read-only, across every one of 159 events, before any change was discussed. ✅ Staged, reversible changes. Data repairs run dry first, then on a handful of records with manual sign-off, before any bulk write. ✅ I use AI-assisted workflows (Claude, Antigravity) to move faster without cutting corners. ✅ I document how it works so the next person can run it without me. Automation and CRM integrations: ☑️ HubSpot: workflows, custom objects, API-driven task and record creation ☑️ ActiveCampaign and Zapier: automation builds, debugging, webhook integrations ☑️ Payment and billing APIs: Chargebee integrations, dunning and revenue-attribution logic ☑️ REST APIs and OAuth: connecting a CRM to whatever else the business runs on Backend and data systems: ☑️ Python: scripts, scrapers, scheduled routines, internal tooling ☑️ Supabase and Google Cloud Platform: error logging, analytics tracking, infrastructure ☑️ Git and GitHub for version control on every build I'm a Mechatronics Engineer based in Mexico, bilingual in English and Spanish, with a Microsoft Azure Fundamentals certification, and comfortable working async across time zones. If you need someone who can get into your tech stack, understand the problem, and build a solution that holds, I'd love to hear about your project.