Summary

Technical builder and forward deployed engineer who takes real customer workflows and turns them into working AI agents in production. I run the full loop end to end: customer discovery and workflow mapping, building and integrating the agent with real systems and APIs, deploying it, debugging it in production, and improving it with evaluation and monitoring loops until it is reliable. I build with modern AI coding tools (Claude Code and custom agents) every day, on a production stack of Python, PostgreSQL/pgvector, Redis, Inngest, Supabase and Vercel. Comfortable with ambiguity, messy legacy systems and high ownership. Systems Engineer; 15+ years directing delivery and engineering teams for enterprise customers, and, since 2024, building AI products hands on with modern AI coding tools.

Experience

AI Engineering & Product Lead · Builder, End to End · North Growth Consulting Oct 2024 · Present

Remote · Early stage GenAI product studio (Anthropic partner program)

Build AI agents and AI products from a blank page to production: Agent Squad (coordinated autonomous agents that automate real operations across dev, QA, PMO and research) and MedicalHubAssist (LLMs and voice transcription for health, recognised by Revista DINERO for its value in the region). I map the workflow, build and integrate the agent with real systems and APIs, deploy it, debug it in production, and improve it with evaluation and monitoring loops (Langfuse). I build with Claude Code and custom agents every day on a stack of Python, TypeScript, PostgreSQL/pgvector, Redis, Inngest, Supabase and Vercel. Under my leadership the company reached Registered Partner status in Anthropic's partner program.

Consulting Manager · Embedded Delivery, Banking · NTT DATA Apr 2021 · Oct 2024

Enterprise customer (Grupo AVAL) · Financial services

Embedded with an enterprise banking customer to turn messy, fragmented workflows into working software. Directed the team that built CrediCore for Banco Popular, a digital credit origination platform on Azure, end to end from discovery to production: I mapped how origination actually worked across fragmented legacy systems (one obsolete, one without vendor support), shipped in phases to get early wins into production fast, and worked directly with the bank's leadership under real ambiguity and partial control. High ownership, direct customer and executive communication.

Digital Delivery & Project Lead · Customer Facing · VASS · GlobalHitss · Tecnocom · earlier Oracle ERP/SCM 2012 · 2021

Pensions · Health · Public Sector · Industrial ERP

Delivered digital products for external customers: modernised a pension fund's digital channels (Porvenir, millions of users; LATAMDIGITAL 2019 award) and a self serve payment settlement portal, and led delivery of a digital commerce portal for a health insurer (Colsanitas). Earlier, as a project manager I ran a project on Oracle's ERP suite, materials logistics module (SCM / supply chain), giving me domain exposure to industrial ERP and supply chain (SCM) workflows in complex, fragmented legacy environments.

Technical

AI & agents: multi agent orchestration, LLMs, RAG, function calling, embeddings, pgvector, prompt engineering, fine tuning, evals and monitoring (Langfuse) · Anthropic API, OpenAI API

AI coding tools: Claude Code, custom agents and skills, Cowork

Build & data: Python (scripting, ETL, automation), reads TypeScript, SQL / PostgreSQL, Redis, API testing (Postman)

Infra & delivery: Supabase, Vercel, Inngest, Hetzner VPS, CI/CD · Agile/Scrum, Jira, Linear · customer discovery, forward deployed delivery

Domain: credit origination, payments settlement, ERP / SCM materials logistics, regulated enterprise · strong professional English (EN/ES)

Summary

Experience

Technical

Education