← Portfolio

Leonel Ulloa

Product Engineer | AI Systems & Automation

Runcorn, UKleonel090810@gmail.comleonelulloa.comgithub.com/leonelulloalinkedin.com/in/leonel-ulloa-ai

Open to: Full Stack Engineer · Product Engineer · AI/Automation Engineer · Founding Engineer · Solutions Engineer

Based in UK · Remote / Hybrid · English & Spanish

Summary

Product and systems builder with hands-on experience designing and shipping production software across SaaS, automation pipelines, business workflows, and internal control systems. Strong in TypeScript, Next.js, React, Supabase/PostgreSQL, workflow automation (n8n), video rendering (ffmpeg), large language model (LLM) integration, API development, and cost-aware product architecture. I use AI tools (Claude Code, ChatGPT) within professional development workflows (VS Code, Git, CI/CD) to build production-grade, reliable systems — not no-code prototypes. I work best on cross-functional problems where product, operations, growth, and technical execution all matter at once.

I am a problem solver by nature — when I encounter a challenge, I study it until I find a solution that is not just functional but improves reliability, reduces cost, and opens new business opportunities. I build systems with production-grade patterns: kill switches, circuit breakers, graceful degradation, and measurement frameworks — because knowing a system works is as important as building it. I think in complete solutions: not the fastest fix, but the one that prevents future problems and creates long-term value.

Selected Experience

Founder / Product Engineer

2025 – Present

PhotoAI Advantage 2.0

AI-powered ad creation SaaS (pre-launch)

  • Built a multi-language SaaS for AI-assisted ad creation using Next.js, React, TypeScript, Supabase, Stripe, Cloudflare R2, Redis, and Inngest
  • Designed and implemented a large product architecture with 215 PostgreSQL tables, 193 API routes, 139 RPC functions, 107 triggers, and 420+ RLS policies
  • Built a 70+ page admin control platform with RBAC, TOTP 2FA, CSRF protection, audit trails, feature flags, promotions, and operational controls
  • Reduced generation cost by separating text rendering from AI image generation, enabling multiple ad variants, resizing, and translation with minimal incremental cost
  • Implemented multi-provider language model orchestration across Gemini, OpenAI, and DeepSeek with fallback logic, reliability safeguards, concurrency limits, and cost guardrails

Tech: Next.js 16, React 19, TypeScript, Supabase, PostgreSQL, Stripe, Cloudflare R2, Upstash Redis, Inngest, Fabric.js, Satori, Puppeteer, Tailwind CSS

Founder / Automation Engineer

2025 – Present

L7AI Content Intelligence

Production automation & orchestration system (live, running daily)

  • Built and operate 27 production n8n workflows that automate the full content lifecycle: research, scoring, script generation, voice synthesis, video rendering, and publishing across 10 social channels in English and Spanish
  • Engineered a short-form video pipeline that produces fully branded videos in 90 seconds at $0.046 per video — replacing 8 hours/week of manual editing with 40 minutes of script review, producing 86 videos/month across two languages
  • Built a custom ffmpeg video renderer (12x faster than browser-based Remotion rendering) with word-level subtitle synchronization, B-roll layering, branding overlays, and progress bar — deployed as a Docker containerized API
  • Implemented production-grade reliability patterns: kill switches (database toggle to halt pipelines instantly), circuit breakers (auto-pause after 5 errors in 5 minutes), rate limiting on paid API calls, graceful degradation (B-roll fallback chain: Pexels → Pixabay → FLUX AI still), and error-only Telegram alerting to prevent alert fatigue
  • Solved infrastructure-level problems including Hetzner IPv4 blocks by building a Cloudflare Worker proxy for ElevenLabs and enabling Docker IPv6 networking, and diagnosed Traefik multi-network routing failures causing 504 timeouts
  • Designed a measurement framework tracking render time, cost per video, success rate, and fallback triggers — all logged to Supabase with real cost validation against Cloudflare R2 file sizes
  • Deployed and maintain self-hosted infrastructure (Hetzner, Docker, Coolify, Traefik) at $43/month with 47 database tables, 14,900+ records, and automated health monitoring via Telegram bot

Tech: n8n, ffmpeg, Docker, Supabase, PostgreSQL, Cloudflare R2, Cloudflare Workers, ElevenLabs, Remotion, Traefik, Postiz, Claude API, OpenAI, Stability AI, FLUX

Founder / Full-Stack Builder

2025 – Present

leonelulloa.com

Bilingual personal brand website (live)

  • Built and deployed a bilingual personal website with 72 AI prompts, 3 ebooks, email-gated downloads, newsletter capture, and analytics
  • Implemented EN/ES routing, searchable prompt library, and lead-capture flow on self-hosted infrastructure

Tech: Next.js, React, TypeScript, Tailwind CSS, Supabase, Docker, Coolify

Technical Skills

Product & Systems

Product architecture, business workflow analysis, admin platforms, internal tools, feature flags, structured workflow design, system reliability, cost-aware system design

Full-Stack

TypeScript, JavaScript, Python (scripting), Next.js, React, Node.js, Tailwind CSS, REST APIs, API integrations, Fabric.js, Satori, Puppeteer

Database & Backend

PostgreSQL, Supabase, RPC functions, Row-Level Security, triggers, schema design, auth flows, rate limiting

Automation & AI

n8n, Zapier, Make, ffmpeg (video rendering pipelines), large language model (LLM) integration (Claude, GPT, Gemini, DeepSeek), AI-powered workflow automation, prompt engineering, structured outputs, TTS with word-level timestamps, image generation, video production pipelines

Infrastructure & Reliability

Docker, Coolify, Traefik, Hetzner, Cloudflare R2, Cloudflare Workers, Redis, GitHub Actions, CI/CD, kill switches, circuit breakers, rate limiting, graceful degradation, error-only alerting, system monitoring, production debugging

Business & Growth

Digital marketing strategy, content strategy, multilingual systems, automation for growth operations, cost/margin analysis

Education

Systems Engineering Technician

University Gerardo Barrios — Computer Engineering

2009 – 2012

Certifications

MicrosoftAzure Fundamentals: Cloud Concepts
MicrosoftAzure Fundamentals: Architecture & Services
MicrosoftAzure Fundamentals: Management & Governance
Google / CourseraDigital Marketing
Google / CourseraE-commerce
Cloud AcademyPython & SQL
Cloud AcademyData Engineering

Languages

Spanish — Native  |  English — Professional working proficiency