Jordan
Gibson
Specialized in voice agents on Retell AI and n8n, scalable LLM job pipelines, and self-hosted Model Context Protocol (MCP) servers.
Engineered for reliability with idempotency, replay protection, cost controls, and automated simulation test suites.
About Me
I am an AI Agent Engineer who designs, deploys, and operates production agentic systems from initial architecture to live operation.
I focus on the operational details demo systems skip: deterministic tool calling, fallback chains, evaluation suites, and cost control.
Authored a 34-tool MCP server allowing AI agents to autonomously manage SaaS operations and support triage.
Constructed safety-gated voice agents using Retell AI, n8n, and multi-tier speech-to-text fallbacks.
Shipped scalable Next.js 16 platforms, tRPC APIs, pg-boss job queues, and low-level C# audio utilities.
Architected a 41,000-line full-stack platform with vision AI metadata generation and dynamic Redis caching.·Brooklyn, NY
Built end-to-end voice interview to print book pipeline coordinating 18 job handlers via pg-boss.·Brooklyn, NY
AI agents must be resilient, safe, and cost-effective. Every workflow I build integrates automated evaluation gates and fallback mechanisms.
Whether creating multi-node voice agents or authoring 34-tool MCP servers for SaaS management, I bridge raw model capability with hardened backend infrastructure.
Selected Work
Founder & AI Engineer·Brooklyn, NY
Engineered a 34-tool MCP server and autonomous Claude Sonnet support triage agent with human-in-the-loop review.
AI Engineer·Brooklyn, NY
Safety-gated voice agent on Retell AI and n8n with emergency screening, 2FA verification, and idempotent routing.
Lead Full-Stack Engineer·Brooklyn, NY
High-end A/V platform with 157 tRPC endpoints, dynamic Redis TTLs, and AI-driven vision metadata extraction.
ML & Backend Engineer·New York, NY
Street-sweeper prediction platform using XGBoost/scikit-learn quantile regression and custom map tile encoding.
Open Source Author·Brooklyn, NY
C#/.NET 8 tray utility interfacing with Windows Core Audio and undocumented COM interfaces for bit-perfect audio.
Let's collaborate on production-ready AI workflows, voice systems, and backend platforms.
Contact
Current base
Typical response turnaround