AI automation engineer
I design, deploy and operate AI automation on my own infrastructure. Then I keep it running.
Built, deployed and operated by me. No one watching them overnight.
Live Case study
Send a photo or PDF of an invoice over a messaging channel. An LLM extracts the tax fields, validates the totals and files them. No manual data entry.
Live Case study
A content pipeline with its own analytics. It publishes on a schedule and measures which decisions actually worked.
Case study
QA and fixes for a dispatch platform in production in Colombia. Every issue filed with its cause, its fix and how to verify it.
Live
A multi-product sales engine over an official WhatsApp BSP. Conversational AI, voice transcription, payment-receipt validation and follow-up windows.
Configuration, not a rewrite.
How a new product launches on the engine
Live
Curates AI news, drafts the post and waits for my approval on Telegram before publishing.
14
sources curated
Programmatic video
Renders vertical videos from a database with Remotion and headless Chrome.
2
YouTube channels, in two languages
Local-first app
How much money will I have on a given date. Pure Dart domain.
182
tests green
You have a system in mind and want it delivered. I scope it, build it, deploy it and hand it over documented, with the automation, the integrations and the infrastructure it needs to run without me.
Ongoing capacity for teams running automation in production: new workflows, integrations, incident triage and the maintenance nobody has time for: expiring tokens, provider API changes, silent failures.
I join your team for a sprint or a quarter and work inside your process: your repo, your board, your review flow. Useful when the automation or AI-integration work is real but doesn’t justify a full-time hire yet.
A review of automation you already have running, with findings backed by evidence: what breaks silently, what has no error path, what will not survive a second client. You get a prioritised report, not a sales pitch.
No public rates. You get a number in my first reply.
Bots handling conversations over an official WhatsApp provider, pipelines publishing on a schedule, workflows filing invoices every day.
Slack, Notion, pull requests and written documentation. UTC-5 overlaps the entire US working day.
AWS EC2 behind nginx, Docker and self-hosted n8n. I don’t need anyone else to deploy what I build.
Usual stack: React, TypeScript, Astro, Node.js, Tailwind CSS, Flutter, n8n, Redis, Remotion, AWS, Figma, Git
Galactic AIMA
I lead product development and run the infrastructure: the self-hosted n8n on AWS EC2 that the whole team’s automations run on, services provisioned with Terraform and reached through SSM instead of open SSH, and the AI systems the agency sells. I also own QA and fixes on a client platform in production.
Independent
Design and development of digital products for clients across Latin America and Spain. Full-stack work with React, Astro and Next.js; interface design in Figma; workflow automation with n8n.
Own projects and collaborations
Visual identities, social media assets and motion graphics. The foundation behind the design judgement I bring to every product.
Institución Universitaria Pascual Bravo
Software engineering fundamentals, databases, networking and systems architecture, plus continuous self-teaching in modern web development and automation.
Tell me what you need running. I reply in writing, with questions, a rough plan and a number.
Async-first. UTC-5, full overlap with US hours.
Or write directly
davidcirortiz06@gmail.com