U.S. Citizen · 100% Remote · AI Product Engineer

I build AI-native products, agents, and automations that ship.

I am Adrian Yepez, a founder-operator and AI systems builder. I build full-stack AI products, agentic workflows, business automations, dashboards, and the infrastructure needed to run them in production.

Hire meDownload resumeRead case studiesView proof
10+AI and business systems built
6live public proof links
Full-stackproduct, AI, infra, QA
RemoteU.S. work authorized

Selected public work

Proof you can click.

These are public-facing systems and demos. The strongest private/internal work is summarized in the case studies below.

AI SaaS

nutricionista.ai

Problem: Nutrition professionals need CRM, patient management, AI productivity, billing, onboarding, and patient workflows in one product.

Built: Built major SaaS surfaces: admin dashboard, patient portal PWA, CRM, appointments, AI modules, integrations hub, SEO directory, and Stripe billing architecture.

Stack: Next.js, React, TypeScript, Supabase, Postgres, Vercel, Stripe architecture

Result: Large working SaaS product surface with 35+ routes, 80+ components, role-based workflows, and AI feature foundations.

View project
Agent Interface

Agent Bubble

Problem: AI agents need a simple public interaction surface that demonstrates real product behavior, not just a prompt box.

Built: Built a public agent/chat product surface connected to the broader agent workflow ecosystem.

Stack: Next.js, AI workflow architecture, Vercel

Result: Public demo available at dm.m8apps.com.

View project
Agent Infrastructure

Titan Agent Fleet

Problem: AI work was fragmented across tools, machines, memory, credentials, browser tasks, and deployment workflows.

Built: Built a distributed agent operating environment with memory/wiki context, model fallback, sub-agent delegation, scheduled jobs, browser automation, and verification rules.

Stack: OpenClaw, LiteLLM, vLLM, Docker, systemd, Caddy, Tailscale, GPG vaults

Result: Reusable AI operations platform for shipping, debugging, deploying, and auditing software/business systems faster.

View project
AI Demo

Agent Arena

Problem: AI-native projects need memorable public demos that show experimentation and product taste.

Built: Built a public AI-agent battle/demo surface with a distinctive concept and clear public link.

Stack: Next.js, Vercel, AI product design

Result: Public AI-native demo available at agentarena.m8apps.com.

View project
Product Build

M8 Mail

Problem: Email workflows need lightweight product interfaces and backend sync foundations.

Built: Built mail tooling with IMAP sync, auth endpoints, frontend app, deployment, and credential remediation after a security incident.

Stack: Fastify, Node.js, IMAP/SMTP, JWT, bcrypt, Expo, Railway, Vercel

Result: Public product surface available at mail.m8apps.com.

View project

Case studies

The work behind the links.

Titan Agent Fleet

Built a distributed AI agent operating environment across Windows, Linux, macOS, and GPU infrastructure. The system combines persistent memory, wiki context, model routing, local fallback, browser automation, scheduled operations, and sub-agent delegation.

  • Designed live-state-first operating rules so agents verify files, services, builds, and endpoints before reporting completion.
  • Routed workloads by machine capability: coordination, GPU/model serving, media/video processing, and browser automation.
  • Integrated cloud model usage with local fallback through LiteLLM/vLLM and OpenAI-compatible endpoints.

WhatsApp Business Agent

Built a WhatsApp automation MVP with webhook intake, inbound classification, contact/profile upsert, message logging, follow-up scheduling, inbox UI, and simulated outbound fallback.

  • Added webhook routes and business inbox screens.
  • Validated with local builds, local smoke tests, production deploy, and production endpoint checks.
  • Designed the system to work in simulated mode when real Meta credentials are not present.

nutricionista.ai Platform

Built product surfaces for an AI-enabled SaaS platform for nutrition professionals: admin workflows, patient portal, CRM, AI modules, integrations, onboarding, and billing architecture.

  • Built 35+ routes and 80+ components across admin, patient, and AI workflows.
  • Used Supabase/Postgres, Next.js, TypeScript, Vercel, and Stripe architecture.
  • Created role-based workflows for sales, support, marketing, and admin operations.

Capabilities

Where I am strongest.

AI product engineeringAgentic workflowsLLM API integrationTool-calling systemsFull-stack TypeScriptNext.js / React / Node.jsSupabase / Postgres / SQLiteVercel / Railway deploymentsDocker / systemd / Caddy / TailscaleLiteLLM / vLLM / local model servingWhatsApp automationProduction QA and smoke testing

Target roles

AI Product Engineer · Agentic Systems Engineer · Founding AI Engineer

I am a U.S. citizen seeking a 100% remote role with a U.S. company. Best fit: startups and teams that need a builder who can own AI product implementation from idea to production.

Email AdrianM8 Apps