Santiago, Chile

Alejandro Ordenes

AI Engineer & Full Stack Developer

I build LLM-based applications and the software behind them: backend, API integrations and production deployment.

Selected work

WhatsApp AI agents

Claude-based conversational agents that run LabNomade's sales channel. They qualify leads, answer questions and hand the team a structured summary of each opportunity.

  • Before each reply, the workflow fetches the lead's data and history from our own API (Next.js + PostgreSQL) and passes it to the model as context.
  • Structured JSON outputs, per-user message limits, repeated-message detection and automatic cut-off when a conversation stalls.
  • Measured cost per message (~USD 0.027) to size pricing and optimizations.
  • Automatic email alert to the team when a lead qualifies.
  • Claude API
  • WhatsApp Cloud API
  • n8n
  • PostgreSQL
  • Next.js

How it works

  1. 1
    WhatsAppThe customer messages the business.
  2. 2
    n8nReceives the message and orchestrates the flow.
  3. 3
    Own REST APIReturns the lead's data and history.
  4. 4
    ClaudeReplies and returns a JSON with the lead's data.
  5. 5
    PostgreSQLLead and conversation are stored.
  6. 6
    WhatsApp + emailReply to the customer, team alert if the lead qualifies.

Multi-tenant client dashboard

Web dashboard where each client reviews their agent's conversations and leads. Custom authentication with PostgreSQL and JWT, migrated off a third-party backend.

  • Next.js
  • TypeScript
  • PostgreSQL
  • JWT

Futures backtesting engine

Python engine that simulates trading strategies over 2 years of 1-minute bars (ES and NQ) against prop-firm risk rules, with incremental data downloads.

  • Python
  • pandas
  • Parquet
StartTrend

StartTrend

Autonomous LLM content pipeline: scrapes TechCrunch and VentureBeat every 30 minutes, summarizes each story with gpt-4o-mini, and every night writes insight articles from the day's summaries as structured JSON with category and sources. Each article gets an illustration generated with the OpenAI Responses API.

  • Java 17
  • Spring Boot
  • MongoDB
  • OpenAI API
  • Jsoup
View code ↗
ArbStrat

ArbStrat

Web app that scans crypto markets for arbitrage opportunities using a cointegrated-pairs strategy, computing statistical metrics in real time.

  • Python
  • Django
  • React
  • PostgreSQL
View code ↗

Experience

  1. Nov 2025 – Present

    AI Engineer & Full Stack Developer · LabNomade

    AI agents and automation for small and mid-sized businesses: requirements, architecture, development, deployment and support.

  2. Sep 2023 – Oct 2025

    Full Stack Developer · Freelance

    REST APIs, dashboards, web scraping pipelines and deployment on Linux servers with Nginx and CI/CD.

  3. Sep 2022 – Jan 2023

    Mobile Developer · BIGEO

    Cross-platform agricultural management app built with React Native.

  4. Jul 2022 – Sep 2022

    Full Stack Developer · Dynae

    Features across the backend (Java, Spring) and frontend (JavaScript) of envia.cl, a Dynae client's platform.

  5. 2021 – 2023

    Computer Programming Analyst · Duoc UC

Let's talk.

Reach me by email or LinkedIn and I'll get back to you shortly.