Rahul B.

Most "AI automations" break the moment real users touch them — no error handling, no retries, no fallback when the API times out or the LLM hallucinates a tool call. I build the kind that doesn't. I'm an AI engineer who builds production-grade AI agents, n8n/Zapier automations, and backend systems for startups and SaaS companies — including hands-on work inside Lovable, Replit, and other AI-native app builders — the kind that run unattended for months, not demos that fall apart after the first client call. ✅ What I build: - AI agents & multi-agent systems (Claude, OpenAI, Gemini) - AI-native apps built in Lovable & Replit - n8n / Zapier workflow automation with proper error handling & retries - RAG pipelines & vector search - MCP servers & custom tool-calling - AI chatbots & voice agents - CRM automation (HubSpot, Salesforce, GoHighLevel, Pipedrive) - API integrations & backend systems (Node.js, Python, FastAPI, PostgreSQL) What sets this apart from a typical automation gig: Most freelancers wire APIs together. I design systems — with proper architecture, monitoring, human-approval steps where they matter, and documentation so your team isn't locked to me forever. ✅ Tech: Claude API, OpenAI/GPT-4o, Gemini, LangChain/LangGraph, n8n, Lovable, Replit, Node.js/TypeScript, Python/FastAPI, PostgreSQL, Supabase, Docker, AWS. Have a workflow that's still manual, an AI feature you want built right, or an existing automation that keeps breaking? Send me a message with what you're trying to solve — I'll tell you honestly whether it's a good fit before we start. 📩 Click "Invite to Job" or Send a Message to discuss your project!