Akshay Pal S.

🏠 I build automated systems for real estate investors and wholesalers that turn property data into qualified seller conversations, booked appointments, acquisitions, and dispositions. 🎯 I’m not just an automation developer who happens to work with real estate. 🚀 I’ve already built and operated a production system for a real estate wholesaling operation that handles the full process: 🏠 Property Data → 🔎 Owner Enrichment → 💬 Seller Outreach → 🤖 AI Conversation → 🎯 Qualification → 📅 Appointment → 🔄 CRM → 👤 Human Handoff Most "AI automation experts" can connect two apps with a webhook and call it a day. I build the systems that those webhooks were supposed to grow into, and then I make them survive contact with real users, real scale, and real edge cases. I'm a freelance automation engineer and AI systems developer with around 7 years in tech and the last few focused entirely on AI, automation, and custom backends. I don't just operate no-code tools. I self-host infrastructure, write the integration and orchestration logic, design the data flow, handle compliance, and own the system end to end, from "here's a messy idea" to "this runs in production and I don't get 2am calls about it." If your project lives in the awkward gap between "a single Zapier zap" and "hire a whole engineering team," that gap is exactly where I work. What I build: AI Voice & SMS Agents. Inbound and outbound lead qualification, booking, and follow-up agents built on Retell AI, Vapi, Claude, and Twilio, wired into CRMs with live call-result capture, transcript logging, and webhook enrichment. I've deployed and maintained agent systems across 200+ sub-accounts. I handle the unglamorous parts too: persona prompt design, human-takeover logic, routing branches, and the "why did the bot say that" debugging that makes or breaks a voice deployment. Custom Automation Backends. Self-hosted n8n on a VPS (Caddy reverse proxy, HTTPS, the works), used for the jobs off-the-shelf automation can't touch: API orchestration, multi-step pipelines, XHR interception where there's no public API, scheduled jobs, and data sync between systems that were never meant to talk to each other. When a platform says "that's not possible," this is usually how I make it possible. Data-Driven Web Apps. Single-page tools backed by real data and real math, not just pretty front-ends. Example: a property underwriting and valuation app built on the ATTOM real estate API with a custom calculation engine supporting multiple deal types and a dual-source valuation model (AVM plus comp-based median). You bring the spreadsheet logic, I turn it into a tool your team actually uses. AI Processing Pipelines. Call transcription and analysis pipelines using Twilio, Whisper, and Claude, complete with usage-based billing, cost tracking, and markup logic so the system pays for itself. The kind of build where the AI part is easy and the billing, error handling, and cost control are where the real engineering lives. GoHighLevel at Scale. Multi-tenant white-label agency architecture, A2P 10DLC registration, webhook and REST integrations, conditional routing, and CRM automation across hundreds of sub-accounts. I know where GHL's bodies are buried and how to work around them. Microservices & Integrations. Small, single-purpose services and webhook endpoints that glue platforms together cleanly, with proper error handling and logging instead of a fragile chain held together by hope. How I actually work: I scope the real problem before quoting, because half of automation jobs are secretly a different job. I flag risks, compliance issues, and bad assumptions up front, even when it's not what you want to hear, because finding them in week one is cheaper than in production. I document well enough that you or your next developer can pick it up without a séance. And I care about what happens after the demo, which is a surprisingly rare trait in this corner of the market. Core stack: n8n (self-hosted) · GoHighLevel · Claude API · Twilio · Retell AI · Vapi · Whisper · REST/webhook integrations · API orchestration · VPS infrastructure (Hostinger, Caddy, HTTPS) · multi-tenant SaaS architecture · real estate data APIs (ATTOM) · A2P 10DLC compliance I also work comfortably with the broader AI and no-code ecosystem (ChatGPT, Gemini, Zapier, Make, Voiceflow, and rapid app builders) when a project calls for speed over custom engineering. The short version: if you need something actually engineered instead of stitched together and quietly praying, that's the whole job for me. Send over what you're trying to build, even if it's still a messy idea. The messy ideas are usually the fun ones.