ANDRES QUINTERO
AI Engineer · Automation and self-hosted infrastructure
- andrestech77@gmail.com
- linkedin.com/in/jaime-andres-quintero-532572240
- github.com/jaimeaq7
- andresquintero.dev
- Valencia, Spain

Profile
I build conversational agents, automate processes and run the infrastructure they live on, end to end. When the data is sensitive I work with local models and a segmented network, so it never leaves the client's perimeter.
Experience
AI and infrastructure developer
Apr 2025 – Jul 2026Legaltech company
- Sole technical profile in the company: development, applied AI, infrastructure and security.
- RAG over Spanish case law and legislation (CENDOJ, BOE) with Vertex AI Search and an own FastAPI backend: hashed keys, signed URLs, sessions and audit trail.
- Communications hub unifying WhatsApp, email and telephony: calls transcribe themselves and arrive with a summary, key dates and actions.
- On-premise server on Proxmox and OPNsense: segmented network and self-hosted stack, removing the recurring SaaS cost.
Founder and operator
2017 – presentOwn subscription business
- More than 400 customers acquired by me: social media acquisition, sales and support over WhatsApp.
- Recurring billing in InvoiceNinja, with automated collection and reconciliation.
- I run it single-handed because it is automated end to end.
Founder and operator
Jan 2016 – Aug 2017Subscription proxy service, own business
- More than 150 paying customers, billed by connected device count.
- AWS infrastructure deployed and operated entirely by me, single-handed.
- Continuity under denial-of-service attacks: address rotation, migration and recovery without cutting customers off.
Founder and owner
2019 – 2021Own restaurant
- Own venue with up to seven employees, built through the pandemic on demand captured beforehand on social media with automations of my own.
- Self-hosted Odoo on Docker for sales, inventory, purchasing and invoicing; cameras installed and operated by me.
Own projects
Private AI infrastructure
In productionOwn server with a dedicated GPU: language, voice and document-search models running locally, behind a segmented network.
- Proxmox
- GPU passthrough
- Local models
- OPNsense
- WireGuard
- HAProxy
Recurring billing and payment validation
In productionRecurring billing and OCR receipt validation: account, approval number and date, with no human in the loop.
- n8n
- OCR
- LLM
- MySQL
- InvoiceNinja
Infrastructure operations agents
In productionAgents running local models that audit, create machines and deploy services. Nothing is applied without human approval, and every action lands in Git.
- Proxmox
- OPNsense
- Local models
- Git
Self-hosted voice agent
In developmentTranscription, model and speech synthesis running locally on an own PBX with a SIP line.
- SIP
- Self-hosted PBX
- STT
- Local LLM
- TTS